Microsoft FYQ4: One of the Strongest Earnings Reports in Multi-Decade History

Posted on July 31, 2025June 30, 2026 by io-fund

Recently, in the Top 15 AI Stocks analysis it was stated “If Nvidia holds the crown in the AI hardware arena, then Microsoft holds the crown in the AI enterprise arena.” Tonight, Microsoft proved why the AI Enterprise crown is rightfully theirs.

Management came out swinging this evening on multiple fronts. First off, the acceleration in Azure and Other Services to 39% up from 35% last quarter was significantly higher than expected, with the Street calling for growth of 33.7%. To grow nearly 40% at this scale is impressive.

Microsoft also revealed its Azure revenue number for the first time of $75 billion for FY2025 (although not entirely surprising as we were modeling for Azure to be hitting $80 billion very soon). From there, the CFO guided for 37% growth in Azure for next quarter – indicating continuing a high growth rate at scale will not be a problem in the near-term (note, H2 is expected to see lower growth than H1).

However, if we look at Commercial RPO, it’s clear something big is going on. Last quarter, we pointed out that Commercial RPO was the one key metric we were watching, stating: “Commercial RPO growth above 30% suggests that Microsoft’s stock could (finally) resume strength again.” At the time, RPO was at $315 billion, up 34% and 33% on a constant currency (CC) basis.

This quarter, Commercial RPO has accelerated to $368 billion, up 37% and 35% on CC basis. Microsoft’s Commercial RPO was in the mid-$100 range in 2022-2023 period to help illustrate how quickly contracted revenue has grown. Wow. We do not typically see such large growth rates on such a large RPO base. It’s almost inconceivable.

A few years back, I described in detail why AI is first and foremost an enterprise technology, specifically calling out Microsoft’s path to $100 billion in AI revenue by 2027. We are seeing this materialize now. Microsoft is putting formidable distance between itself and best-of-breed cloud players. To illustrate, stocks like Confluent are down 27% after hours following the loss of a large customer.

In addition to the key metrics stated above, management carries a sense of confidence when analysts question the ROI on capex. And when Mark Zuckerburg boasted about building a gigawatt-plus cluster called Prometheus next year, Satya made sure to lead his introduction by saying “We stood up more than 2 gigawatts of new capacity over the past 12 months alone.” You’ll find more commentary on this below.

Revenue – Azure reported as standalone segment for first time

Revenue was up $76.4 billion for growth of 18% or 17% in constant currency. This is up from last quarter with growth of 13% or 15% in constant currency and beat consensus of $73.83 billion. For the fiscal year ending in June, the company reported revenue of $281.7 billion, up 15%.

Azure revenue was reported as a standalone metric for the first time, being stripped out of “Azure and Other Services.” The company stated Azure saw $75 billion in revenue or growth of 34%. For comp purposes, the original segment grew 39% up from 33% / 35% on CC basis last quarter.

Below, you can see the visible acceleration in overall revenue

Below you can see that 39% is the highest growth rate we’ve seen in some time for Azure and Other Services:

Looking forward, management guided for revenue of $75.25B at the midpoint, beating consensus of $74.15B. This would represent growth of 14.7%.

According to the CEO, Microsoft is ahead of other hyperscalers in speed of data center buildouts: “We continue to lead the AI infrastructure wave and took share every quarter this year. We opened new DCs across 6 continents and now have over 400 data centers across 70 regions, more than any other cloud provider. There is a lot of talk in the industry about building the first gigawatt and multi-gigawatt data centers. We stood up more than 2 gigawatts of new capacity over the past 12 months alone. And we continue to scale our own data center capacity faster than any other competitor.”

Revenue segments – Cloud has highest growth rates since 2022

Cloud reported some of its highest growth in three years. The CEO stated: “Through software optimizations alone, we are delivering 90% more tokens for the same GPU compared to a year ago” as well as “

Microsoft Cloud was up 27% and up 25% on CC basis for revenue of $46.7B. This marks the highest quarterly growth rate since CY2022
Gross margin was 70% up 100 basis points from 69% last quarter
Productivity and other Businesses was $33.1 billion, up 16% and 14% on CC basis.
Intelligent Cloud was up 26% and up 25% on CC basis for revenue of $29.9 billion. This was the highest growth rate since CY2022
More Personal Computing was up $13.5B for growth of 9%

Commercial Bookings Surpasses $100 Billion for the first time

To help support the case for future growth, both commercial bookings and commercial RPO came in surprisingly strong.

The CFO stated that for the first time commercial bookings surpassed the $100 billion mark, increasing 30% on CC basis. Commercial RPO increased to $368 billion, up 35% on CC basis with 35% recognized in revenue in the next 12 months.

Additional key metrics: 500 trillion tokens processed last year; 800M AI Product Users

Azure is always the main metric looked at, yet we should pause and share a few more important key metrics in this banner report.

Copilot apps have surpassed 100 million monthly active users across commercial and consumer.
Across broader AI features, there are over 800 million monthly active users.
Foundry Agent Service is now being used by 14,000 customers to build agents.
80% of Fortune 500 use Foundry, processing 500 trillion tokens, up 7X YoY.
Microsoft Fabric is a data and analytics platform for AI workloads, with revenue up 55% year-over-year and over 25,000 customers. According to management: “It's the fastest-growing database product in our history.”
There are 20 million GitHub Copilot users. GitHub Copilot enterprise customers increased 75% quarter-over-quarter and 90% of the Fortune 100 use GitHub Copilot.

Margins & Earnings

EPS of $3.65 beat consensus estimates of $3.38.

Gross margin was 68.5% up from 68.1% last quarter for gross profit of $52.4B.
Operating margin of 44.9% was up from 44% for operating profits of $34.3B.
Net margin was 35.6% up from 34.9% last quarter for net profits of $27.2B.

Cash flows & capex raised to eye-watering $30B per quarter

Operating cash flow of $42.6B was up 15% YoY
Free cash flow of $25.6B was up 10% YoY
Capex of $24.2 billion was up 27% YoY with management guiding for capex of $30 billion next quarter.

Earnings Call Q&A:

Capex Spend Correlates to $368B in RPO:

Every Big Tech company will be asked about ROI on capex spending, and the CFO handled the question quite well, stating: “when you think about the full year comments I've made on CapEx as well as the Q1 guidance of over $30 billion, you first have to ground yourself in the fact that we have $368 billion of contracted backlog we need to deliver, not just across Azure but across the breadth of the Microsoft Cloud.

So in terms of feeling good about the ROI and the growth rates and the correlation, I feel very good that the spend that we're making is correlated to basically contracted on the books business that we need to deliver and we need the teams to execute at their very best to get the capacity in place as quickly and effectively as they can.

And so when you look, and we've talked about the growth rate [of capex] will decline year-over-year, but at its core, our investments, particularly in short-lived assets like servers, GPUs, CPUs, networking storage, is just really correlated to the backlog we see and the curve of demand. And I talked about, my gosh, in January and said I thought we'd be in better supply demand shape by June. And now I'm saying I hope I'm in better shape by December.”

Conclusion:

This was an earnings report for the ages – simply because the Commercial RPO is massive, and Microsoft is proving they can grow at a scale we haven't seen yet in AI software. Earlier today, I had stated on Bloomberg that Microsoft could see $40 billion in AI revenuesometime in 2026 – which is a massive number, but what's most important is the rapid ascent in reaching that number.

If you zoom-out, a few years back I've made the case that Microsoft could see as much as $100 billion in AI revenue by 2027 and then I upped it to $200 billion by 2028. Should we see this ballpark figure, it would mark a rapid ascent hard to fathom a few years back. This earnings report is a step in the right direction to meet that mark.

Please note: The I/O Fund conducts research and draws conclusions for the Fund’s positions. We then share that information with our readers. This is not a guarantee of a stock’s performance. Please consult your personal financial advisor before buying any stock in the companies mentioned in this analysis.

Recommended Reading:

Meta Reports Large Q2 Beat, Ad Impressions Growth Rebounds

Posted on July 31, 2025June 30, 2026 by io-fund

Meta reported a large beat on both the top and bottom line in Q2, with revenue nearly $3 billion ahead of estimates at $47.52 billion and EPS of $7.14 more than 21% ahead of consensus. Ad impressions growth meaningfully rebounded from the mid single-digits to the double-digits in Q2.

For Q3, Meta also set the bar quite high by guiding for $49 billion in revenue, pointing to a consecutive quarter with >20% growth. For context, Meta was expected to report <15% growth in both Q2 and Q3.

In the bigger picture, Meta gave a line of sight to $100 billion in capex in 2026 as it works to build out its Prometheus accelerator cluster, which Mark Zuckerberg stated is expected to be the first 1+ GW scale cluster to come online next year. Combined with accelerating expenditures, there will possibly be some pressure on operating margin next year.

Revenue Growth Returns to 20% Range

Meta reported 21.6% YoY growth to $47.52 billion in Q2, nine points ahead of its guidance for 12.6% growth to $44 billion at midpoint. Advertising revenue was robust, growing 21.5% in the quarter, a more than five point acceleration from 16.2% growth in Q1.

According to management, the beat was driven by AI unlocking greater efficiency and gains in their ad system: “This quarter, we expanded our new AI-powered recommendation model for ads to new surfaces and improved its performance by using more signals and longer context. It's driven roughly 5% more ad conversions on Instagram and 3% on Facebook.” They also pointed toward advancements in the recommendation system helping to lead to a 5% increase in time spent on Facebook and 6% on Instagram.

In addition, Meta is expanding its tools for smaller advertisers “with a meaningful percent of our ad revenue now coming from campaigns using one of our generative AI features.”

For Q3, Meta guided for a consecutive quarter of >20% growth, seeing revenue between $47.5 to $50.5 billion, or 20.7% YoY at midpoint. This includes a 1% growth tailwind from FX.

Meta did not provide formal guidance for Q4, though it did state that it expects YoY growth to be slower than Q3 as it laps a stronger comp in Q4 2024.

Advertising Key Metrics – APAC Outperforms in Impressions, ARPP Rises Nearly $2 YoY

Ad impressions meaningfully rebounded from 5% growth in Q1 to 11% in Q2, though to confirm this quarter as a possible inflection back to double-digit growth, Meta would need to either maintain this growth or accelerate from here in Q3.

US & Canada impressions rebounded to 9% growth, up from just 4% in Q1.
Europe impressions growth was 6%, up slightly from 5% in Q1. 
APAC outperformed with impressions growth of 16%, up from 9% in Q1.
Rest of World impressions growth was 7%, rebounding from 1% in Q1.

However, ad pricing decelerated slightly, down from 10% in Q1 to 9% in Q2. Meta said this slight deceleration was from stronger impression growth, though pricing benefitted from “increased advertiser demand, largely driven by improved ad performance.”

US & Canada pricing increased 11% YoY, decelerating from 14% in Q1. 
Europe pricing increased 17%, accelerating from 9% in Q1. 
APAC pricing increased 2%, compared to 3% in Q1. 
Rest of World pricing increased 15%, compared to 17% in Q1.

Overall, ad revenue growth was the strongest in Europe and Rest of World at 24% and 23%, respectively, while US & Canada and APAC grew 21% and 18% YoY.

Meta’s Family of Apps daily average people (DAP) increased 6.4% YoY to 3.48 billion, and average revenue per person (ARPP) rebounded to 14.8% YoY to $13.65. This is a nearly $2 increase from Q2 last year, and closing in on Q4 2024’s record at $14.25.

Meta is working to expand its monetization pathways across its Family of Apps, rolling out video and image ads to Threads and ads to statuses and channels on WhatsApp. However, Meta acknowledged that Threads will not be a meaningful driver, while WhatsApp ads earn lower average prices that Facebook/Instagram as usage is skewed towards lower priced markets.

Margins Improve, but Watch 2026 Expenses

Gross margin held steady with Q1, and operating margin improved sequentially. However, Meta made a handful of key comments about 2026 expenses that could pressure margins down the line.

Gross margin was 82.1%, flat QoQ and up 0.8 points YoY.
Operating margin was 43.0%, up 1.7 points QoQ and 5 points YoY.
Net margin was 38.6%, down 0.7 points QoQ but up 4.1 points YoY.

Meta provided some initial commentary for 2026 expenditures, signaling an acceleration in expenses, with growth likely far outpacing revenue growth. Meta said that 2025 total expenses will increase 20-24% YoY, and 2026 total expenses are likely to increase at a higher rate.

Rising infrastructure costs are the primary driver from a “sharp acceleration in depreciation expense growth and higher operating costs as we continue to scale up” data center infrastructure, followed by employee compensation. Management also singled out energy costs, finance leases and increased spending on cloud services to meet capacity needs as other factors behind higher operating expenses.

However, revenue growth for 2026 is expected to be just 14% to $215 billion, though given the magnitude of Q2’s beat and Q3’s guide (at nearly $6 billion above estimates combined), this is likely too low. Assuming revenue comes in closer to $230 billion, with an 82% gross margin and a 27-30% increase in expenses, operating margin may fall from ~39-40% towards 34-36%.

EPS

Meta recorded a massive EPS beat in Q2, reporting 38.4% growth to $7.14 versus estimates for 14% to $5.90. This was driven by operating leverage and Meta’s large revenue beat.

Heading into the report, Meta was expected to see negative earnings growth in both Q3 and Q4, but given that EPS beat estimates by 21%, it’s likely that these estimates move higher in the coming days/weeks considering the upbeat revenue guidance for Q3 and expenditure guidance remaining relatively unchanged.

For FY25, Meta was expected to report 8.5% growth to $25.90, but given the nearly $1.30 beat in Q3, FY25 EPS is likely to be revised to $27+, or at least 13% YoY.

Cash Flows and Balance Sheet

Operating cash flow margin dipped sequentially, though FCF took a larger hit due to rising capex. Following Meta’s large-scale move to acqui-hire Scale AI for nearly $15 billion, cash and equivalents have shrunk sharply, with analysts questioning how management plans to fund increasing capex come 2026.

Operating cash flow was $25.56 billion, for a 53.8% margin. This contracted from a 56.8% margin in Q1 but was up from a 49.6% margin in the year ago quarter.
Free cash flow was $8.55 billion, down more than (21%) YoY as capex rose more than 100% YoY to $17.01 billion. FCF margin was 18%, down from a 24.4% margin in Q1 and a 27.9% margin in the year ago quarter.
Cash, equivalents and marketable securities totaled $47.07 billion, down from $70.23 billion last quarter. Debt remained steady at $28.83 billion.

Earnings Q&A: 2026 Capex in Focus, Nearing $100B

Meta’s earnings call focused primarily on its heightened expenditures for next year and 2026 capex approaching $100 billion, as well as the monetization pathways for these investments. Management reaffirmed a goal to build out multiple GW of data center capacity in a quest to reach superintelligence, under the belief that it will “improve every aspect of what we do”. However, management also was clear in stating that genAI will not be a meaningful revenue driver this year or next.

Management stated that while the “infrastructure planning process remains highly dynamic, we currently expect another year of similarly significant CapEx dollar growth in 2026 as we continue aggressively pursuing opportunities to bring additional capacity online to meet the needs of our AI efforts and business operations.” Given that 2025’s capex is currently guided to increase ~$30 billion YoY at midpoint, this suggests Meta is targeting $100 billion in capex next year.

CFO Susan Li provided more clarity on the driving factors of this capex increase, saying that Meta expects “a greater mix of our CapEx to be in shorter-lived assets in 2025 and '26 than it has been in prior years,” or higher spending in servers, networking, and data centers to continue building out AI training and inference capacity.

Given the fact that cash and equivalents have shrunk rather dramatically, analysts questioned about how Meta will finance this capex, as it will strain cash flows. Li said that Meta does expect to finance “some large share” of capex itself, though it is exploring other avenues to co-develop data centers with financial partners. She added that some models will “will attract significant external financing to support large-scale data center projects,” giving them some degree of flexibility in their cash spending.

Analysts also questioned Meta about the ROI on this capex, considering how its aggressive push into the metaverse backfired. Management said that “on the core AI side, we continue to see strong ROI,” such as the visible impacts to ad conversions and impressions. On the genAI side, Meta said it is “clearly much, much earlier on the return curve and we don't expect that the genAI work is going to be a meaningful driver of revenue this year or next year,” though the company remains very optimistic about long-term monetization pathways. However, the question here is how much will Meta pour into capex before genAI becomes a meaningful driver – including 2026’s estimate, Meta’s three-year capex is already approaching $200B+.

Damien Robbins, Equity Analyst at I/O Fund contributed to this analysis.

Recommended Reading:

Positions Report – July 2025

Posted on July 30, 2025June 30, 2026 by io-fund

I/O Fund Portfolio and Game Plan

Going into this year, we advocated for a defensive posture due to various risks in the broad market, as well as within valuations. We outlined these risks each week on our webinars going back as far as October of 2024. As a result, we maintained a very high cash allocation and quickly jumped into our hedge at the first sign of support breaking in January and February.

As a result, we remained largely unscathed in the April drop and deployed nearly all our cash into the lows. In fact, we issued 22 buy alerts between March 3rd and April 7, with 12 of those buy alerts happening between April 4th – 7th. We do not know many out there that were buying these lows as aggressively as we were.

Since the April lows, we have maintained our game plan – reposition defensively into the bounce and let the market tell us what the next move will be. While we did not expect this bounce to go this high, and began positioning defensively too early, we are still adhering to the same gameplan – if the next pullback is a messy/3-wave drop, we will position aggressively for the next trend higher. If instead we see an aggressive/5-wave drop, we will patiently wait for lower levels to position.

This game plan is best expressed by the two counts presented in the chart below:

Green – We are completing the first leg in a large degree 5th wave. The 2nd leg should take the shape of a 3-wave decline that first breaks below 6200 and then 6105. It will then find support on 5900 – 5500 and must hold 5345. We will then turn higher in a more aggressive 5-wave uptrend to new all-time highs.
Red – The final 5th wave completed in February of 2025. The bounce off the April low is a large corrective rally within a larger uptrend. The next drop will be an aggressive/5-wave drop that ultimately breaks through 5345.

Note how the current rally is pushing higher with a strong deceleration in momentum and volume. This is a common occurrence when we are close to a trend change. Furthermore, the RSI is still hovering below bear market resistance for the S&P 500, signaling that a return to the level of momentum that we tend to see in strong and lasting bull trends is not yet present.

One additional point is noting how many key markets are currently not participating in the S&P 500’s recent rally to new highs. Key markets that have either led the bull market or are more sensitive to the economy did not confirm the S&P 500’s push to new highs last week.

While divergences can correct, and the trend continue, nearly every turning point – tops and bottoms – shows divergences like the ones shown above. It implies that the higher we go, the less key markets are participating.

While the odds are quite high that we should see some type of reversal soon, we still favor the green count based on the strength of the A.I. market, which appears to be ramping, as expected. The stock market is one of the greatest predictors and leading indicators. For this reason, we will remain open, and risk focused, as we wait for the market to signal the “all-clear.”

Current I/O Fund Portfolio

Our current portfolio is back into a defensive posture. Our game plan is to rebalance our portfolio for the second half of 2025. We are first raising cash around the highs and will issue our buys on the next drop.

As of now, we are heavily tilted toward A.I. hardware, which we plan to maintain. However, we intend to increase our exposure to A.I. software, as well as energy in the coming rebalance.

Regarding crypto, the easy gains are over. We are on the lookout for the top to the current bull cycle in the coming weeks or positioning for one more high into the $200,000 region. Our exposure to crypto will depend on how the next drawdown shapes up. We go into much greater detail in the individual positions below.

Astera Labs (ALAB)

ALAB has been the stock with the most bullish potential since we began investing in it in 2024. This lines up with the fundamental outlook as we discussed in The I/O Fund’s Top 15 Stocks for Q3 2025, which is why it is such a large position. There are two general interpretations of the price data that we are tracking:

Green – The 2024 uptrend off the low is a large 5-wave pattern. This was followed by a deep 3-wave retrace that made a higher low. We now completed another 5-wave uptrend off the 2025 low, followed by a sideways consolidation for wave 2.
We called out the importance of the $100 – $85 region, as many institutions were active in this region. The breakout move suggests this was accumulation. So, the next drop needs to be 3-waves down, and hold above the $85–$100 region to suggest an imminent breakout.
Blue – The recent push higher was a 5th wave – aka, bull trap. The coming pullback will break through $85, finding support in the $60s–$70s. We must hold $59 if we see this deeper pullback. This will end wave 2, as we set up for the larger 3rd wave breakout.

Bitcoin (BTCUSD)

The weekly chart on Bitcoin has us definitively in the final 5th wave of the bull cycle that started in 2022. Note that price went vertical in 2024 on max volume and momentum. This is the 3rd wave, which is almost always the most exciting part of a trend.

The primary question is when did the 3rd wave end? If it was in mid-2024, which is represented by the red count, then we should see a final push into the $125,000 – $135,000 region. If the 3rd wave ended in late 2024, which is represented by the green count, then we can see the 5th wave take us well over $200,000 before ending.

The internals on the weekly chart of Bitcoin are also at a key inflection point. Note that the RSI and the Detrend Oscillator are at the same levels that have marked meaningful reversals in Bitcoin’s trend. This is also happening on decelerating volume. In other words, the higher price goes, the less interest we are seeing from buyers.

If the more bullish green count is in play, all of these internal signals will have to be broken to the upside while momentum and volume will need to expand from here.

Another concern worth mentioning is the current cycle in play with Bitcoin. The below chart shows Bitcoin trending higher into a cluster of large cycles. What matters with cycles is how we trend into them – more times than not, they indicate a reversal of the trend heading into their window.

While the cycle work that I do is secondary to what price does, this chart further supports the precarious spot Bitcoin is in. Ideally, we would see these cycles pointing to a relative low, or period of consolidation before a bigger break out. However, until the above concerns resolve, we will maintain a defensive posture.

On a smaller time frame, the below chart outlines the potential paths that the price action can take. As mentioned above, we will need to see the internals on the weekly chart improve, as well as supports hold into the current cluster of cycles.

Green – The swing higher is incomplete and should get into the mid $120K to lower $130K range. Once completed, we should see a 3-wave pullback that holds over $105,000. If this breaks, the risk will increase.
The final support for the bullish count will be $93,200 – $86,000. Though the risk will be elevated below $105,000, if we can hold these above support regions, the setup to +$200K will remain intact.
Red – This scenario was stated above. We are in the final 5th wave of the multi-year bull cycle, which is targeting $125,000 – $135,000. We will then see a larger pullback that will break through $105,000, $93,200, and finally $86,000. If these supports break, it will increase the odds that a renewed bear cycle is underway, with general targets in the $70,000 – $40,000 region.

Nvidia (NVDA)

Nvidia has been a difficult stock to map since the June high in 2024. The reason for this can be seen below. Since, we have seen a series of three wave patterns in all directions, while price is only moderately higher than the June 2024 peak.

There are two counts that I’m tracking with NVDA:

Green – It is obvious that the bounce off the April low is taking the shape of a 3-wave pattern. So, this would be the end of wave 1 of a large degree ending diagonal pattern. We would see a 3-wave retrace soon, which holds $116 – $111. We will then turn higher toward the $300 – $350 range in the following months.
Blue – This count also has us in an ending diagonal, but on a smaller degree. In other words, we are already in the 3rd wave, which should be ending soon. The 4th wave drop should hold over $155 – $140 and then turn higher toward $192 – $216. This would complete the final 5th wave in the NVDA uptrend that started in 2022, and we should see a multi-month draw down that follows. If this does happen, it will likely set up a generational buying opportunity.

Bloom Energy (BE)

BE quickly became one of our largest positions on the recent breakout to new highs. There are currently two scenarios that best describe the current price action:

Green – We have completed a leading diagonal for wave 1. This was followed by a rare and shallow 2nd wave. Today’s breakout was the start of wave 3, which should continue into next week. Any weakness has to hold $26.80–$28.50 or this count will be invalidated. The targets for this move are $66 – $108.
Blue – This breakout will fail under $26.80, setting up a drawdown to the $24 – $20 region for a larger 2nd wave. We have to hold $18 for any further bullish pattern to continue.

Coinbase (COIN)

We started a position in COIN in late 2024 around the lows. Since then, we have been tracing what appears to be the last large swing in a very large diagonal pattern. This is a 5-wave pattern that has large moves in both directions, causing significant overlaps within the pattern. We are in the 5th wave of this large pattern, and there are two scenarios that best define the price actions:

Green – I think COIN is going to put in a top before Bitcoin. It needs to hold $370 and then turn higher toward $445 – $476. This will complete a multi-year leading diagonal for wave 1. Wave 2 will take months to a year to play out and will target $149 – $82. This is where the generational buying opportunity lies. It was a good run. We are sitting on sizable gains, and we don’t need to be greedy.
Red – Either we break below $370, or we make another high, as suggested above. We’ll then drop back into the $324 – $290 region, preferably. However, this drop can go as low as $222 and still be valid. As long as $222 holds, we can turn higher in 5-waves toward the $500 – $800 range.

Chainlink (LINKUSD)

We have been holding off on adding to LINKUSD until we get more clarity in the pattern. Since the December 2024 top, Chainlink went from one of the clearer charts that we track to one of the more confusing.

What cannot be ignored is the very large and clear 5-wave pattern that started off the 2023 low that pushed into the late 2023 high. This large degree pattern strongly suggests a higher resolution, which is why we are hanging onto our sizable LINKUSD allocation.

Based on the price action, the below counts represent the most likely outcomes:

Green – The April low was the end of a large corrective bounce. Off this low, we got a nice 5-wave pattern, which further confirms a new uptrend is starting. We have also broken through the downtrend angle as we pushed toward the key $18 region. For this count to be in play, I want to see a continuation higher, while holding over $18. The lowest I can give this count is the $16 region. As long as these levels hold, and we keep pushing higher, then we are targeting the $35 – $54 region.
Red – We are still in the large 4th wave. We will break down below $18, $16, then $11.75, which will setup a final swing lower toward the $8 – $7 region.

Ethereum (ETHUSD)

It’s hard to believe but Ethereum still has not broken above its 2021 high, even though it has risen quite a lot from the 2022 low. The pattern has taken the form of a messy and overlapping bounce, which means that there are two possible scenarios in play:

Green – We are in a small 4th wave that should hold $3036. If held, we will turn higher toward 4120 – $4675. This will complete a large A-wave, which will be followed by a corrective B-wave. The likely targets for this correction will be $2795 – $2137. However, it must hold $1850. We’ll then turn higher toward $8000 – $10,000 in a final blow off.
Red – After this next swing higher (or break of $3036), we will start a new bear cycle. We don’t want to own ETH during this.

Taiwan Semiconductor (TSM)

TSM is following a similar path as Nvidia:

Blue – We are completing the 3rd wave of a small ending diagonal. This would mean that we are due for a pullback that holds over $187. This will be followed by one more swing into the $264 – $315 region.
Green – We are tracing out a large ending diagonal pattern. Instead of completing the 3rd wave, we are completing the 1st wave. This will be followed by a larger correction that can break $187 but must hold $153.

Super Micro (SMCI)

Green – SMCI is tracing a very large 5-wave pattern from the 2008 lows. The -85% drop in 2024 completed the 4th wave correction, as we are now setting up for the 5th wave swing.
For this to be in play, any weakness from where we are should see a drop into $42 – $35 region and hold over $31.50. This will set us up for a larger breakout that is targeting $107 – $181.
Red – We are still in the larger 4th wave correction. The bounce since the November 2024 low has been a corrective bounce, making a lower high, and we are setting up for the final drop to $13 – $9. If we break below $31.50, this count will become the most likely outcome.

Credo (CRDO)

Credo is in the middle of the current Blackwell supply chain, while also having a relatively attractive valuation. It was also a primary target for our buys in March and April. We decided to step aside due to its elevated China exposure, which the market is rightfully looking past. Expect this to be a bigger position, as we strategically layer back into it at key levels.

Green – We are in the final 5th wave of a leading diagonal pattern. The bounce off the April low was the first swing, which should be followed by a 3 wave retrace. This retrace will break below $93 to signal it is underway and should ideally hold $75 – $67; however, it can go as low as $46 and remain valid. Once this correction is over, we should see an aggressive push to new highs, with tentative targets in the $119 – $173 region.
It's also worth noting that there is evidence that institutions are very active in this stock between $88 – $93. If this level holds, and we see a strong break to the upside, it will negate the expected drop. It is rare to see a stock move as far as CRDO has in such a short amount of time with minimal pullbacks. However, its positioning within the Blackwell supply chain should be considered.
Red – We already completed the diagonal pattern in the current swing higher. The coming correction will break $46, to confirm this scenario, and could take months to play out.

Advanced Micro Device (AMD)

Since the 2021 top, AMD has been in a messy consolidation to a very large degree. I think that we are in a very large B wave, which will either end this year or into next year. The below counts represent these scenarios:

Blue – We are completing the final swing in a 3-wave bounce off the April low. Note how we are pushing up into a very heavy supply zone between $160 – $175, and we are doing so on weakening volume and momentum. We should see a 3-wave drop That holds over $101 – $96 before setting up for the next larger swing higher.
Green – We had a tiny correction and setting up for a continuation of the vertical bounce off the April lows. If we can breakout over $160 and then $175, we should see a push into the target zone, which starts at $205.

Solana (SOLUSD)

We continue to get confirmation that Solana could be setting up for a large swing to the $400 – $600 region. For one, we have what can be counted as a 5-wave move off the April low. This was followed by a 3-wave correction that has made a higher low. We are now establishing a momentum position into the constructive price action. Below represents the general counts we are tracking:

Green – We are in a minor dip wave that should target $176 – $152. It has to hold $152. We will then turn higher toward $242, which is where we start taking gains (sell half). We will then need to see either another pullback that makes a higher low, or a vertical breakout through $242 on heavy volume, for confirmation.
Blue – This count has the push off the late June low as a corrective bounce within a larger correction. We will break through $152 in a more aggressive, and direct fashion, then head toward $125 – $111. We must hold $111 if this count plays out, or it opens the door to $60.

Oklo (OKLO)

Oklo is the first stock that we have included from our Discovery Tier. It is positioned within the A.I. energy space, which we believe, and have been saying, is a more immediate need for A.I. stocks.

Green – We have a very large 5-wave pattern off the 2024 low. After a 3-wave pullback that made a higher low, we now have another 5 waves higher. The bullish breakout scenario says that we need to see another dip before an explosion higher. This count suggests that we got an unusually small dip and are setting up for an imminent breakout over $77, preferably on expanding volume and momentum. Any weakness before a breakout would have to be a 3-wave pattern that holds over $55.50, or this count will get invalidated.
Blue – This count has us making a double top. We should see a 5-wave drop from here that breaks through $55.50. The targets for this drop would be between $45 – $30 and must hold $25.

The new Discovery tier offers fast-paced research on new stock ideas the I/O Fund is interested in, with technical setups and comprehensive deep-dive analysis. Be the first to know what exciting new tech, AI and energy stocks the I/O Fund is tracking here. here.

Meta (META)

META is competing a clear 5-wave push off the April low. Note how we are making new highs on weakening volume and momentum. There are two paths that the price action supports, as of now.

Green – We will see a 3-wave correction that holds over $582 – $534. We will then need to see a more direct push off this low, which will setup a large breakout.
Red – We are still in a large degree correction. The current bounce will give way to a 5-wave drop that break below $534. This will setup a target below the April low.

Applovin (APP)

Red – APP is in a large degree uptrend. The drop in February was a correction within this larger uptrend, which completed into the April lows of 2025. We then saw the start of the next swing higher into the June high, which is being followed by another minor correction that is ultimately targeting the $270 region. Once price moves below $324, it will confirm this scenario. We must hold $240 if this scenario plays out, or we will be targeting the $150s in a continuation of the larger correction that started in February. If the $240s hold on any further weakness, we should see a setup for a breakout move that is targeting the $1000s for the next leg higher.
Green – APP will hold $324 and instead see a breakout over $393 – $429. This will set us up for a larger breakout toward the overhead target.

Coherent (COHR)

Blue – Coherent appears to be tracing out a large degree ending diagonal pattern. The April low completed wave 4, which would put us in the final 5th wave swing to new highs. Note how the current bounce has not made a new high. This will simplify COHR, as the next drop has to be a 3-wave pattern, and it must hold over $54. If this happens, the next swing will be a more direct and aggressive pattern that is targeting $136 – $196.
Green – We had a very shallow B wave and are setting up for a 3rd wave move to new highs. This move should be vertical, ideally occurring on a gap and accompanied by expanding volume. The overhead targets are the same at the blue count.

Mara Holdings (MARA)

Green – We are completing a minor correction within a larger uptrend. We must hold $15 on any further weakness. If we do, the next move higher will be around $23. We will then break through $23 and trend toward $29 – $42.
Red – This bounce off the April low is a correction within a larger downtrend. We will either break through $15 imminently or fail on the next bounce under $23.

Dell (DELL)

Based on the obvious 3-wave drop from the 2024 high, there are only two scenarios that fit the current price action:

Green – We are completing the first leg higher in a very large 5th wave. Note the current uptrend is pushing higher on decelerating volume and momentum. This implies that we are setting up for some type of pullback. We should see a 3-wave drop soon that ideally holds $106. However, this drop can go as low at $81 and still maintain this count. The 5th wave target would be around $180 – $200, and potentially higher depending on how shallow the coming retrace is.
Red – We are in a large degree and complex correction from the 2024 top. The current bounce is a corrective bounce within this larger uptrend and should see one more drop below the April low. The next drop will be a 5-wave pattern that breaks through $77 for confirmation.

Potential New Stocks

We have been reducing risk on the current push higher, raising cash and locking in real gains. When we see a correction, the below names will be considered for inclusion into our portfolio.

Oracle (ORCL)

Green – I believe that we are setting up for a decent pullback. The signals are plentiful, and once we move below $228, we will be in that correction. It must be a 3-wave pullback that holds: $210, $187, $172, $165. Then we will turn higher toward $350 -$475.
Blue – We are at all-time highs after a vertical move off the lows. We already had the 3rd wave breakout and are in an ending diagonal. So, we will correct, but hold $228, then head toward $290 – $305 in a 5th wave.

Vertiv (VRT)

Green – VRT is in a large degree 5-wave uptrend. The recent drop was a correction within this larger uptrend, as we are now setting up for the final swing higher. Note that price is flashing multiple sell signals – the detrend oscillator is at the same amplitude where previous corrections started, momentum continues to fade with volume as price pushes higher. We should see a drop back to $93 – $75, and it has to be a 3-wave drop. If any further weakness holds $65, we should be setting up for the larger breakout to new highs.
Red – We are still in the large degree correction. The current bounce off the April lows is a corrective bounce within this larger drawdown. The next drop will be an aggressive 5-wave drop and break through $65. The final targets will be between $53 – $42.

Broadcom (AVGO)

Green – Broadcom has been tracing a very large diagonal pattern since the COVID low in 2020. This count has us completing the first leg higher in the 5th wave. Note how momentum and volume continue to fade the higher we go. We should see a pullback to the $240 region, at minimum. The ideal targets for this correction are between $218 – $180. Once this pullback is over, we should continue to new highs with general targets between $335 – $482.
Red – The bounce off the April low is a large bounce within a larger correction. We will break through $180, which will shift the odds in favor of this count. The targets for this drop will be $139 – $127, and it will complete a large correction within an even larger uptrend.

Palantir (PLTR)

Green – We are in an ending diagonal, which should lead to a swift move back to the $121 – $107 region. This will hold and then turn higher toward the $180 – $215 region. This will complete the 5th wave in a very large 33rd wave.
Red – This move off the April low resembles a 3-wave pattern. It is a B wave and should lead to a 5-wave drop that breaks through $107, and then $96 – $80. The final support will be $66 – $48, and it will complete a very large 4th wave correction within a larger uptrend.

Cloudflare (NET)

NET is a stock that we owned in the past and sold for nice gains. Now that the inference market is heating up, we believe it is worth revisiting. However, the current valuation and stock pattern suggest waiting for better prices.

Green – We have completed a very large 1st wave in a new bullish uptrend. The drop into the April low was wave 2, and we are now in the early stages of the 3rd wave. We need to see a minor pullback into the $148 – $130 region, and it has to hold $119.
Red – We are still in the larger 2nd wave. The bounce off the April lows is a correction within the drawdown, and we should see a 5-wave drop that breaks $119. This will set up a final target in the $98 – $67 region.

There are two more stocks that we intend to add in the coming pullback that are in our Discovery Tier. Join us on Monday, August 18th at 4:30 pm, for our first ever Discovery Webinar! We will go over the technical setups in each stock within this tier, including the two that we intend to add.

To subscribe to Discovery with 30% off, please click here to email usclick here to email us or email premium@io-fund.com and mention code DISCOVERY30.

Recommended Reading:

Google Stock Clears Major Hurdle, Yet One Serious Concern Remains

Posted on July 27, 2025June 30, 2026 by io-fund

In this post, we take a deeper look at Google's earnings results for Q2 and share our thoughts on:

Earnings results for Google Search and updates on the battle between Gemini and ChatGPT
Google Cloud’s results, and what the increase to capex means in competitive context
Threats to Google’s ecosystem—the one serious concern that remains for Google Stock
Who benefits from Google’s $85 billion capex and the AI growth stock we like better than Alphabet

On May 7th of this year, during the Department of Justice’s antitrust trial against Google, Apple's Senior Vice President of Services, Eddy Cue, revealed in his testimony that Safari searches declined for the first time in 22 years. Mr. Cue attributed the decline to the rising popularity of AI-powered search and assistants such as OpenAI’s ChatGTP, Perplexity AI, xAI’s Grok, and Anthropic’s Claude. On the topic of the eventual transition from traditional search to AI-powered alternatives, Cue added, “There’s enough money now, enough large players, that I don’t see how it doesn’t happen.”

In response, shares of Alphabet fell 9%, erasing an estimated $140 billion in market value.

Since then, investors have been awaiting Alphabet’s Q2 earnings to provide clarity around the issue—and it has, at least for Google’s Search business. But other questions remain in terms of one area Google is losing major real estate, which we discuss below.

Search reported 12% YoY revenue growth of $54.2 billion, helping drive Alphabet’s total revenue to $96.4 billion, up 14% YoY. While OpenAI’s ChatGPT, Perplexity’s new Comet browser, the Department of Justice, and others deserve continued monitoring, Google’s resilience in this space is impressive.

On the Q2 earnings call, Alphabet stated that Google is successfully integrating AI Overviews into its core search experience. With 2 billion monthly users across 200 countries, AI Overviews contribute 10%+ additional queries for the searches in which they appear. For Google, this represents a value add and engagement driver, not a disruption.

Google’s Gemini and ChatGPT are battling to become the AI assistant of choice for consumers. To compare, Gemini has reached 450 million MAUs, up 100 million from March of this year, with daily requests surging 50% over Q1. But according to Google’s own data shown in court back in March of this year, ChatGPT had an estimated 600 million MAUs, and likely many more by now. Gemini is gaining ground, though still lags behind ChatGPT.

Alphabet also reported on its launch of AI Mode—a new option displaying as the first tab in Google’s main search menu, next to All, News, Videos, Images, and the rest—in the U.S. and India. The company said AI Mode has already amassed 100 million MAUs.

Capex Now 40% Higher Since the Start of 2025

While Google Cloud revenue reaccelerated four points to 32% to $13.6 billion—a welcome beat considering Q1 had seen a 2-point sequential deceleration —the big news on Cloud is the reported $10 billion increase to capex, now up to $85 billion. This is 40% higher than the $60 billion capex estimate coming into the year, signaling Alphabet's willingness to spend on AI to drive accelerating cloud growth.

Similar to Oracle stock, this surging capex is weighing heavily on free cash flow, and will likely continue to pressure free cash flow in the upcoming quarters. Operating cash flow rose just 4% YoY to $27.7 billion in Q2, yet free cash flow declined (61%) YoY to $5.3 billion. While TTM free cash flow is still up 10% YoY to $66.7 billion, the market is forward looking, and this capex hike implies TTM free cash flow will continue to trend much lower over the next few quarters.

It is understandable why Alphabet is willing to increase its capex, as it is beginning to witness increasing operating leverage in Google Cloud and seeing AI usage proliferate across its core Search business. For example, AI Overviews now serves 2 billion users and processes 980 trillion tokens monthly, doubling since May, a scale which helps explain the explosive compute costs driving Google’s massive investment in Cloud.

Cloud operating margins have expanded significantly, up more than nine points from 11.3% to 20.7%, as operating income rose 141% YoY to $2.8 billion. Prioritizing AI investments in an effort to drive Cloud growth of >30% could see the segment reach a $120 billion run rate by mid-2028, more than double its $54 billion run rate this quarter. Should operating margins begin to expand towards the high-20% to 30% level by mid-2028, versus AWS in the high-30% range, Cloud could generate $30 billion in annual operating income, or more than 3x from today.

This represents more than one-quarter of current company-wide operating income, or a potential massive driver of profitability in the future — the point here is that Alphabet is essentially sacrificing some FCF growth in the near-term and spending heavily in an aim to reap the rewards of a much larger, fast-growing Cloud segment in the future.

Another market reality adding pressure to Google’s Cloud and AI initiatives is the fact that Microsoft Azure has already set a high bar and is leading the race to monetization by a substantial margin. Whereas Google processes huge volumes of tokens—980 trillion, a figure that has doubled since May of this year—Microsoft’s deeply established enterprise relationships with hundreds of millions of users has unlocked user and revenue growth.

Microsoft’s fiscal Q3 report helped cement the company as the strongest AI player in the hyperscale crowd due to its focus and dominance across enterprise software offerings and deep AI integrations aided by its partnership with OpenAI.

I/O Fund covered Microsoft’s impressive Q3 2025 earnings and Azure’s outperformance on May 15th, which you can read for free here: Microsoft Stock Surges After Q3 2025 Earnings: What Separates Azure from AWS, Google CloudMicrosoft Stock Surges After Q3 2025 Earnings: What Separates Azure from AWS, Google Cloud

One Serious Concern Remains for Google’s Stock

The intense regulatory pressure on Google to open its ecosystem, even as leading AI assistants are knocking on the door, is a lot for Google to overcome. The DOJ’s proposed remedy—Alphabet divesting itself of the Chrome browser—comes as rivals like Perplexity, OpenAI, and Anthropic are developing assistants that can integrate with other Android apps. The pressure on Google to open its ecosystem is mounting, and the competition is already finding their way in.

Primarily, it’s the loss of Google Search being the default search engine on Samsung devices that poses a larger threat to the company, in our opinion. This is because any loss of real estate – especially on mobile – during a time when there are fierce rivals only creates more headwinds for the stock.

As was reported in early June, Samsung is entering a deal with Perplexity to become the default AI search engine on their devices. According to sources such as Bloomberg, Samsung plans to preload the Perplexity AI assistant and app onto its smartphones, with its AI search built directly into Samsung’s Internet Browser. There could also be AI-powered operating systems built later down the line for multi-agent platforms. This shift could happen as soon as 2026.

Here are a few of the key developments that could lead to Google Search losing market share on mobile, specifically, in the next 1-2 years:

OpenAI, Perplexity, and Anthropic are developing assistants that can integrate with other Android apps, undermining Gemini usage on devices.
OpenAI has expressed interest in acquiring the Chrome browser to boost already strong usage of ChatGPT, while Perplexity recently launched its own agentic browser, Comet.
On April 1st, Google signed a new, non-exclusive agreement with Samsung that includes no restrictions on the smartphone manufacturer loading alternative search products.
Anthropic recently introduced the ability of its Claude assistant to integrate with Google Workspace, and OpenAI will soon introduce its own assistant that can connect with Android apps and Google Workspace.
While Google is also in talks with Apple to integrate its Gemini assistant with Siri, Google Search is unlikely to enjoy the "default" status that it has enjoyed for many years on iPhones, given that Apple is also in talks with other AI rivals.

The takeaway is that Google’s ability to leverage one of its key distribution channels during the generative AI revolution is already compromised. The real threat kicks in when Android smartphone users have the choice to opt for third-party digital assistants over Gemini—subsequently undermining Google’s advertising revenue growth.

One Stock that Will Benefit from Google’s $85 Billion Capex

As I pointed out in the Bloomberg interview above, the best way to position is with stocks that directly benefit from Big Tech’s increase in capex.

While Alphabet is still a very solid business with the major advantages outlined in this analysis, its $85 billion capex commitment could require years to generate meaningful returns in the Cloud. Therefore, we prefer to invest in companies that benefit strongly from capex over those that spend heavily on capex.

Keep in mind that some companies will benefit from capex across all of Big Tech – therefore, the $85 billion from Google is just the start to the tailwinds for specific AI hardware stocks. Investors should factor in there is similar spend from Big Tech: Alphabet, Amazon, Meta, Microsoft and Apple.

Below, my firm outlines a direct beneficiary of Google’s enormous capex spending – which was increased 40% this year alone and was raised $10 billion in the most recent earnings report. After all, one stock’s loss is another stock’s gain.

For investors hungry for more near-term growth, we recommend Broadcom (AVGO) as it directly supplies tensor processing units (TPUs) to Google.

Alphabet is a primary custom silicon chips (ASICs) customer for Broadcom, with HSBC estimating the giant will account for nearly three-fifths of Broadcom’s ASICs shipments in its fiscal 2026.

With Alphabet’s newest TPU version, Ironwood, expected to carry a premium $13,000 ASP, versus Broadcom’s other customers at ~$5,000 , this could drive a 128% YoY increase in ASICs revenue to $28.4 billion (not including AI networking). Alphabet’s capex increase in Q2 further cements this story into Broadcom’s AI growth thesis for next year.

Why Big Tech Is Chasing Cheaper Inference

For the providers in the AI ecosystem, monetizing GPUs depends on inference, and thus revenue becomes a function of GPUs and tokens and profits become a function of cost. Nvidia’s Blackwell offers a massive leap in performance and can train models such as Meta’s Llama 3.1 405B in as little as 27 minutes, yet the cost advantages offered by custom silicon can translate into higher margins in the long run from lower inference serving costs.

For example, Google recently announced that its upcoming seventh-gen TPU Ironwood is its “most performant and scalable custom AI accelerator to date, and the first designed specifically for inference.” Ironwood comes in two sizes, a 256 and a 9,216 chip configuration, with the larger size offering up to 42.5 exaflops of performance.

Google adds that Ironwood offers 2x the performance per watt as last-year’s generation Trillium, with 6x more HBM and 4.5x the HBM bandwidth. This allows it to deliver more capacity per watt at a time when power is a primary constraint, and provide customers with more cost-effective AI workloads.

This is exactly what Broadcom sees arising from this inference growth curve, as CEO Hock Tan asserted that the company has quite a bit of visibility into “increased deployment of XPUs next year, much more than we originally thought and hand-in-hand with it, of course, more and more networking.” The necessity of networking in larger clusters means demand is likely to remain robust even given custom silicon will not keep pace with Nvidia’s merchant sales into the hundreds of billions.

Higher-than-expected deployments of custom silicon combined with strong demand for networking should provide robust tailwinds for AI revenue growth beyond 2026. Broadcom currently has enough visibility to place possible demand acceleration for 2H 2026 on the table, and this could easily persist through 2027 and beyond should inference demand flourish and as the path to 1 million accelerator clusters materializes.

Assuming Broadcom can maintain another 60% YoY growth in FY27 on stronger demand and potential conversion of its 4 current prospects, AI revenue would close in on $50 billion, or up to 60% share of revenue. Even if growth then slows to 30% YoY in FY28, Broadcom would still be more than doubling its AI revenue to $65 billion in just three years.

Broadcom Reports 170% YoY Growth in AI Networking

Broadcom has cemented itself in second place in AI revenue as it closes in on $20 billion this fiscal year in AI revenue — with a line of sight toward $30 billion by the end of fiscal 2026. AI revenue accounted for more than 50% of Semiconductor revenue for two quarters in a row and nearly 32% of total revenue in Q2.

AI semiconductor revenue rose 46% YoY to $4.4 billion, in line with management’s guidance. Although this was a deceleration from 77% YoY growth in Q1, Broadcom forecast $5.1 billion in AI revenue in Q3, pointing to a rebound to 60% YoY growth – marking ten consecutive quarters of growth.

In the current quarter, the 46% AI semiconductor growth was driven by networking, which was up 170% YoY and represented 40% of AI revenue. In the opening remarks, the CEO stated the following regarding this outsized growth: “As a standard-based open protocol, Ethernet enables one single fabric for both scale out and scale up and remains the preferred choice by our hyperscale customers. Our networking portfolio of Tomahawk switches, Jericho routers and NICs is what's driving our success within AI clusters in hyperscalers.”

Q3’s guidance was ahead of some analyst expectations for $4.9 billion in AI revenue in the quarter, ticking higher as Google’s TPU v7p (Ironwood) begins to ramp. Q3 would also mark the largest sequential growth in over a year on a dollar basis, at ~$700 million.

Additionally, analysts look to already be penciling in further strength in Q4, with Bernstein’s Stacy Rasgon suggesting that Broadcom could be eyeing $5.8 billion in AI revenue in Q4 assuming it sustains 60% YoY growth. Given that Broadcom’s 1H revenue was up more than 57% YoY, this seems a reasonable assumption, especially considering management is eyeing near 60% growth in FY26.

More importantly, AI’s strength is masking persisting softness in non-AI revenue, which could continue to be pressured due to Broadcom’s high consumer exposure. Broadcom noted that non-AI revenue “is close to the bottom” but it “has been relatively slow to recover” with revenue down (5%) YoY to $4 billion in Q2.

Despite this weakness extending into Q3 with revenue expected to be flat QoQ at $4 billion, semiconductor revenue is accelerating – growth accelerated from 11% to nearly 17% in Q2, with the $9.1 billion semiconductor revenue guide pointing to an acceleration to nearly 25% growth in Q3.

Should non-AI revenue soon find the bottom and begin to recover, this will provide support for continued Semiconductor growth. However, any persisting weakness in non-AI stemming from this elevated consumer and Apple exposure that AI revenue must absorb presents a real risk that investors should keep in mind through the rest of the year. Broadcom is also one of the more exposed semiconductor companies to China with tariffs, with more than $10 billion in revenue from the nation in fiscal 2024.

Broadcom Stock to See Lift from AI Inference

Broadcom is aiming to capture growing inference tailwinds, with management explaining that the recent surge in inference demand is driving increased confidence in their FY26 AI revenue growth rate.

CEO Hock Tan said that Broadcom’s hyperscale clients are “doubling down on inference in order to monetize their platforms,” and as a result, he expects Broadcom could “actually see an acceleration of XPU demand into the back half of 2026 to meet urgent demand for inference on top of the demand we have indicated from training.” This new dynamic is what is driving Tan’s confidence in stronger growth in FY26, saying that he now anticipates the “fiscal 2025 growth rate of AI semiconductor revenue to sustain into fiscal 2026.”

This commentary plus potential demand acceleration in 2H 26 suggests that Broadcom has visibility into $30 billion AI revenue potential next year. Broadcom has not provided a full FY25 AI revenue guide yet, but it is on track to deliver approximately $19 to $20 billion in AI revenue in FY25, up ~60% YoY assuming 60% growth to $5.9 billion in Q4.

Maintaining 60% growth through FY26 would project AI revenue to $30 to $32 billion. This trajectory indicates Broadcom is likely driving AI revenue ahead of expectations over the next four to six quarters, with Morgan Stanley saying that $26 to $30 billion in AI revenue is “higher than what is in Street models.” Evercore is modeling 58% AI revenue growth in FY25 and 50% in FY26, implying $28.9 billion.

Valuation is Too High for Broadcom

Broadcom is a key beneficiary of Google’s capex yet the stock is richly valued.

On the top-line, Broadcom trades at nearly 22x forward revenue, a 5% premium to Nvidia’s 21x multiple. AVGO stock was at a 14% premium heading into Q2’s earnings. This is also 85% higher than Broadcom’s 5-year average 11.8x forward revenue multiple.

On the bottom line, Broadcom trades at 43.8x forward earnings, an 8% premium to Nvidia. Broadcom has strong margins – 65% adjusted operating margin and 52% adjusted net margin – driving strong EPS growth, at a 25% expected CAGR through FY27; however, the custom silicon ramp presents some headwinds to gross margin as it grows its mix share.

To find out key levels the I/O Fund plans to buy Broadcom next, consider joining our 1-hour webinar on Thursday at 4:30 pm EST – plus the one thing the CEO stated that all investors must hear in this article “This Stock is Set to Surge from Inference Demand” reserved for premium members only.

For 2025, the I/O Fund has worked to identify key Nvidia suppliers with Blackwell on deck to ramp significantly, sharing our in-depth research on the AI networking stack. Sign up to join our upcoming webinar, held every Thursday at 4:30 pm EST. Our cumulative 5-year results would place us #2 if we were a hedge fund and #5 if we were an ETF. Learn moreLearn more

Recommended Reading:

Amphenol Reports Strong Q2 on 133% Datacom Growth

Posted on July 25, 2025June 30, 2026 by io-fund

Amphenol reported a strong Q2 beat with revenue of $5.65 billion coming in far ahead of its guidance for $4.9 billion to $5.0 billion, riding a second consecutive quarter of 133% YoY growth in IT datacom. Its Communications segment also helped drive stronger operating margin expansion, which flowed through to the bottom line with an impressive 28% GAAP EPS beat versus estimates.

However, Amphenol stated that they shipped substantially more than expected in datacom in Q2, essentially pulling forward some demand into the quarter and reducing Q3’s sales. Analysts placed this pull-forward effect at ~$150 million, which management confirmed as an accurate rough estimate for the impact. This is likely the primary factor behind Q3’s guidance for a QoQ decline in revenue, which is not typical for Amphenol.

Q2 Revenue Beats Estimates by 12%, yet Q3 Guided for Sequential Decline

Amphenol reported 56.6% YoY growth in revenue to $5.65 billion, far ahead of its guidance for 36-39% YoY growth to $4.9-5.0 billion. Organic revenue growth was 41% YoY, accelerating from 33% YoY in Q1 and 20% YoY in Q4.

For Q3, Amphenol guided for $5.4 billion to $5.5 billion in revenue, up 34-36% YoY, but down (3.5%) sequentially. This goes against typical seasonality for Amphenol, with Q1 usually only quarter to see a QoQ decline. It also suggests that Q2 may shape up to be the company’s peak growth quarter with this 21 point guided deceleration.

To note, consensus estimates have continued to rise since our prior analysis in June, with Q3 revenue up $110 million since then to $5.24 billion, and Q4 up $140 million to $5.39 billion. Given the magnitude of Q2’s beat at $610 million and Q3’s guidance $210 million above consensus, it’s likely that FY25 revenue estimates will move higher by ~$1 billion, to closer to $21.5 billion. This would project FY25 revenue growth at 41.4% YoY, or 7 points faster than current estimates of 34.3%.

For a more detailed look at APH and its revenue opportunities from Nvidia’s Blackwell, refer to the June analysis here: https://io-fund.com/premium/amphenol-reports-134-growth-in-datacom-it-segment https://io-fund.com/premium/amphenol-reports-134-growth-in-datacom-it-segment

Key Segment Breakdown

Amphenol has three primary reportable segments, Harsh Environment Solutions, Communications, and Interconnect and Sensor Systems, though it also breaks down growth by end market (discussed next). These three all serve many of the same end markets, such as industrial, auto, and datacom, so the end market breakdown provides a clearer view of what’s driving growth.

Communications Solutions revenue increased 101.4% YoY to $2.91 billion, accelerating from 93% YoY growth in Q1. Organic growth in the segment was 78% YoY, accelerating from 73% in Q1.

Harsh Environment Solutions revenue increased 38.2% YoY to $1.45 billion, maintaining growth in the 38% range. Organic growth accelerated to 18% YoY from 8% in Q1.

Interconnect and Sensors Systems revenue increased 15.7% YoY to $1.30 billion, accelerating from 5% in Q1; this also marked the segment’s fastest growth in more than six quarters. Organic growth accelerated eight points sequentially to 14% YoY.

The chart below shows each segment’s quarterly growth since Q3 2023, with the pace of acceleration in Communications clearly visible. As a result, Communications now accounts for more than 51% of Amphenol’s revenue, up from 40% a year ago.

End Market Breakdown

IT datacom revenue remained robust, rising 29% from Q1 to more than $2 billion in revenue.

IT Datacom grew 133% YoY and organic, accounting for 36% of revenue in Q2. Revenue rose 29% QoQ, outpacing Amphenol’s guidance for a high single-digit QoQ increase, driven mainly by AI related products as management stated that roughly 2/3 of IT datacom’s growth on a YoY and QoQ basis were driven by AI. For Q3, growth is expected to decline mid to high single-digits QoQ, as Amphenol mentioned that they “shipped substantially more than expected, including some modest portion of third quarter demand.”
Industrial grew 25% YoY and 12% organic to 19% of revenue in Q2, with growth in alternative energy, medical, factory automation, and more. Amphenol said the segment’s growth was much better than expected on a QoQ basis, while Q3 is expected to decline slightly due to seasonality.
Automotive grew 10% YoY and 8% organic to 14% of revenue. Q3 sales are expected to be slightly lower QoQ, with customers planning for typical summer shutdowns.
Communications networks grew 143% YoY and 16% organic, with the YoY print stemming primarily from Andrews acquisition. For Q3, revenue is expected to be flat QoQ.
comm nwk operators; 30% QoQ in Q2; sales to remain flat QoQ
Defense grew 25% YoY and 18% organic to 9% of revenue, with growth across most segments within the market. Amphenol also closed its $300 million acquisition of RF and microwave supplier Narda-MITEQ in the quarter, with the company having annual revenue of $120 million.
Mobile devices grew 14% YoY and organic to 6% of revenue, as strength in laptops was offset by declines in tablets. For Q3, growth is expected to increase high single digit QoQ.
Commercial aerospace grew 50% YoY but just 8% organic to 5% of revenue, benefiting from the CIT acquisition from last year. Q3 revenue is expected to be up low single-digits QoQ.

Orders and Book-to-Bill:

Orders rose approximately 4% sequentially to $5.52 billion in Q2, a new record, though growth slowed from 58% YoY to 36% YoY.

Book to bill was 0.98:1, the first time this ratio has dipped below 1:1 since Q4 2023. Management noted that book to bill was slightly stronger in defense market, and was modestly lower in IT datacom due to shipping ahead of expectations in Q2.

Margins

Driven by increasing contribution from its Communications segment, margins expanded quite substantially on a sequential basis.

Gross margin expanded more than 2 points sequentially to 36.3%.
Operating margin expanded 3.8 points sequentially to 25.1%. On a YoY basis, operating margin has expanded 5.7 points.
Adjusted operating margin expanded 2.1 points sequentially and 4.3 points YoY to 25.6%.
Net margin expanded 4 points sequentially to 19.3%.
However, adjusted net margin expanded only 1.6 points sequentially to 18.6%, as Amphenol recorded some SBC-related tax impacts.

On a segment view, Communications is the clear driver of this expansion, with operating margin surpassing 30%. However, all segments saw operating margin improve sequentially.

Communications operating margin was 30.6% in Q2, up 3.2 points sequentially and 6.3 points YoY. Should the segment maintain this operating margin or see it expand further in the low-30% range, Amphenol could continue to see operating and net margins expanding, fueling strong EPS growth.
Harsh Environment Solutions operating margin was 25.2% in Q2, up 0.7 points sequentially and 0.4 points YoY.
Interconnect and Sensor Systems operating margin was 19.5%, up 1.4 points sequentially and 1.3 points YoY.

Amphenol did not provide specific margin guidance for Q3, but management did state in the earnings Q&A while discussing operating margins that Q3 “really reflects another strong quarter of profitability, kind of essentially at the same levels on slightly lower revenue guidance.”

EPS

Amphenol reported a strong EPS beat in the quarter, as margins outperformed — adjusted EPS beat expectations by nearly 21% while GAAP EPS beat by more than 28%.

Q2 GAAP EPS of $0.86 increased 110% YoY, beating estimates for $0.67.
Q2 adjusted EPS of $0.81 increased 84.1% YoY, beating estimates for $0.67. This growth was 37 points ahead of guidance for nearly 48% growth, and a more than 26 point acceleration from Q1.

For Q3, Amphenol guided for adjusted EPS of $0.77 to $0.79, or YoY growth of 54-58%. This also was ahead of estimates for $0.71 in the quarter. Growth is expected to decelerate rather sharply into Q4.

Cash Flows and Balance Sheet

Cash flow margins strengthened significantly in the quarter, though debt remains a concern as Amphenol recently priced more than $1.3 billion in senior notes. Inventories and accounts receivable have been rising over the last two quarters, suggesting demand remains high.

Operating cash flow increased more than 113% YoY to $1.42 billion, for a 25.1% margin. This increased from a 15.9% margin in Q1 and an 18.4% margin a year ago.
Free cash flow was $1.12 billion, for a 19.8% margin. This increased from a 12.1% margin in Q1 and a 14.6% margin a year ago.
Inventories have jumped substantially over the past two quarters, rising from $2.55 billion in Q4 to $2.91 billion in Q1 and now to $3.13 billion in Q2.
Accounts receivable have followed inventories, rising from $3.29 billion in Q4 to $3.92 billion in Q1 and now to $4.27 billion in Q2. Combined with rising inventories and strong order growth, this supports strong growth heading into Q3.
Cash and equivalents totaled $3.23 billion, while debt was $8.06 billion. Amphenol also recently priced two tranches of senior notes in June, a 3.125% €600 million trance due in 2032 and a separate 4.375% $750 million trance due in 2028. Proceeds will go towards debt repayment and general corporate purposes.

Earnings Call Q&A:

Amphenol spent much of the Q&A discussion talking about their ability to execute and outperform expectations in Q2, though there were a handful of questions about the durability of AI demand, capex and margins.

Evercore’s Amit Daryanani questioned about AI infrastructure growth moving forward in light of Amphenol’s slight Q3 pull in: “I realize there was about $150 million of shipment in June versus September that you talked about. But really away from that, can you spend some time talking about how do you think about the durability of growth on the AI infrastructure side as you go out over the next few quarters?”

Amphenol gave a very vague response to this question:

“If you think about our second quarter, we originally guided the quarter to be at the high end, $5 billion in sales, and we ultimately achieved $650 million more than that. And on the IT datacom side, we outperformed very, very significantly. And you gave a side to that, and that's, I think, a good rough estimate of how much we kind of shipped of what would be Q3 demand. And if you factor that in, you certainly don't see a peakiness to the performance.”

Management followed up by saying that there is “continued momentum" in IT datacom, but every quarter cannot be expected to grow 133% YoY. Over the near and medium-term, management remains confident in growing IT datacom, hinting that they are winning content in next-gen GPU systems regardless of whether fiber optics or high-speed interconnects are used. They did not provide any insights into which next-gen systems, or what bill-of-material content could be, leaving an open question on growth come 2026.

Building on this, Amphenol highlighted that Q3 capex will likely be higher than its typical 3% to 4% of revenue range, as it works to support this IT datacom growth. Management backed this up by saying that they “have an expectation and a confidence that we are continuing to gain momentum in this space, winning programs, getting visibility from customers for their future plans, which create incremental opportunities for us,” driving this need for higher capex potentially over the next few quarters. This could potentially weigh on free cash flow if operating cash flow normalizes back to the mid to high-teens from 25% in Q1.

Additionally, Amphenol discussed long-term conversion margin targets given Q2’s outperformance on operating margin. CFO Craig Lampo said that the company has “long targeted the 25% conversion margin in the last couple of years, and really meaningfully exceeded that benchmark” in Q2 due to strong organic growth and a shift to selling higher-tech, higher-margin products, such as those for AI applications.

He added that the company is expecting some normalization of margins moving into 2026 as they scale costs in line with higher revenues, but they believe that “conversion margin will continue to remain higher, meaningfully higher than kind of that 25% conversion target that we've historically had.” Over the longer-term, Lampo raised this view, saying that “close to 30% would be our target kind of moving forward.” This opens the door for EPS growth to possibly strengthen over the next few years, with current growth in FY26 and FY27 only two to three points above revenue growth at 13% and 11% respectively.

Conclusion

Amphenol’s Q2 was quite strong, with revenue accelerating sharply on robust AI-driven datacom demand. However, Amphenol acknowledged that this outperformance pulled forward a portion of datacom demand from Q3 into Q2, and overall revenue was guided to decline slightly sequentially. Despite this, margins have strengthened, driven by Communications segment, and are expected to remain strong again in Q3 with EPS guided nearly 10% above consensus.

The I/O Fund owns AI networking stocks that are linked to Nvidia and custom silicon projects such as Amazon’s $100B capex including Trainium. We share our portfolio with Pro and Advanced Members. Advanced Members also receive real-time trade alerts, entries, exits and trade plans in our weekly webinars. Take advantage of a limited-time offer for $75 off Pro or $100 off Advanced. Learn more hereNvidia and custom silicon projects such as Amazon’s $100B capex including Trainium. We share our portfolio with Pro and Advanced Members. Advanced Members also receive real-time trade alerts, entries, exits and trade plans in our weekly webinars. Take advantage of a limited-time offer for $75 off Pro or $100 off Advanced. Learn more here

Damien Robbins, Equity Analyst for the I/O Fund, contributed to this analysis.

Recommended Reading:

Taiwan Semiconductor Q2 Earnings: FY25 Guidance Raised on Strong AI Demand

Posted on July 23, 2025June 30, 2026 by io-fund

Taiwan Semiconductor reported earnings last week, providing the first glimpse into the AI semiconductor industry in the second quarter. Riding strong AI chip demand, TSMC boosted its full year revenue growth guidance, yet there are still some lingering doubts about the chipmaker’s growth in Q4.

HPC revenue continues to accelerate, reaching a record at $18 billion in Q2. Net profit reached a record in Q2 as margins outperformed. With that said, TSMC signaled significant margin pressures in Q3 from FX and ramping overseas fabs.

TSMC boosts FY revenue by 5 points

Q2 revenue increased 44.4% YoY and 17.8% QoQ in USD to $30.07 billion, well above TSMC’s guidance for $28.4 billion to $29.2 billion in revenue. This was driven by strong demand for AI accelerators built on TSMC’s 3nm and 5nm nodes.

For Q3, TSMC guided revenue of $31.8 billion to $33 billion. At midpoint of $32.4 billion, this represents YoY growth of 37.8% and QoQ growth of 7.8%. FX is playing a role here in this high-37% guide, with TSMC noting that revenue growth in NT$ is expected to be negatively impacted by 6.6 points based on current exchange rates, or ~31.2% YoY.

For the full-year, TSMC boosted its revenue growth forecast from mid-20% YoY to close to 30% YoY, driven by robust AI and HPC demand.

As it stands, TSMC’s growth in the first three quarters is far above the close to 30% guide, suggesting Q4 could see growth in the single digits. Management said they remain more conservative on Q4 at the moment. This is because although Q4 is typically seasonally strong for consumer electronics and smartphones, there is risk that tariffs put a damper on growth.

Management added that they have not seen any changes in customer behavior, but they are well aware of uncertainties from tariffs on consumer and price-sensitive end-markets. They added that Chinese rebate programs are stimulating near-term demand upside, but this is expected to phase out rather quickly and only create a mild recovery in non-AI demand this year.

HPC revenue rises 14% QoQ and up 19.2% CC

TSMC continues to benefit from robust AI accelerator demand, with HPC now accounting for three-fifths of the chipmaker’s revenue. TSMC stated that HPC revenue rose 14% QoQ in NT$ in Q2, though this increase was more pronounced on a US$ basis due to FX. TSMC’s revenue is recognized in US$, so every 1% appreciation of the NT$ adversely impacts NT$ reported revenue by ~1%.

Therefore, on a US$ basis, HPC revenue rose 19.2% QoQ to ~$18 billion, or up nearly $3 billion from Q1 in constant currency — its largest growth on record. The pace of acceleration in HPC revenue has been astonishingly quick, as the segment is now 2.5x larger than it was two years ago at $6.9 billion. HPC’s share of revenue also has increased 8 points YoY to 60%.

Smartphone revenue increased 7% QoQ to account for 27% of revenue in Q2, notably the strongest sequential increase for the segment since 2022.

IoT revenue increased 14% sequentially to account for 5% of revenue, while automotive was flat and also accounted for 5% of revenue. DCE increased 30% sequentially to account for 1%, while other segments rose 14% sequentially to 2% of revenue.

Revenue by Technology

TSMC’s advanced nodes – 3nm, 5nm, 7nm, and soon, 2nm – contribute the majority of revenue at 74% in Q2, fueled by AI accelerators and smartphones. 3nm ticked back up to 24% of revenue in Q2, while 5nm held flat at 36% of revenue, supported by Nvidia’s Blackwell GPUs. 7nm saw its contribution shrink one point to 14%, while mature nodes also shrunk one point to 26%.

TSMC shared some more details about its upcoming advanced nodes, stating that it remained on track for volume production on its 2nm node beginning in the second half of 2025, with a ramp profile similar to the 3nm node. Management also stated that they “expect the number of new tape-outs for 2nm technology in the first 2 years to be higher than both 3nm and 5nm,” driven by HPC and smartphone.

For the A16 node (1.6nm), TSMC said that it remains on track for volume production in the second half of 2026, believing this node will be best suited for “HPC applications with complex signal routes and dense power delivery networks.”

Margins guided to decline due to FX headwinds in Q3

Despite some FX headwinds to gross margin, TSMC’s operating margin outperformed, driving profit to a record level in Q2. However, Q3’s guide showed increasing FX headwinds and sharper sequential impact on margins.

Gross margin was 58.6% in Q2, at the high end of the guided range for 57% to 59%. Gross margin declined sequentially from 58.8%, with a 2.2 point headwind from FX and a 1 point headwind ramping overseas fabs offset by higher capacity utilization.
Operating margin was 49.6%, increasing 1.1 points sequentially from operating leverage, above guidance for 47% to 49%. Operating margin was up more than 7 points YoY.

Net margin was 42.7%, down slightly from 43.1% in the prior quarter but up nearly 6 points YoY.

For Q3, margins will decline due to FX:

TSMC guided gross margin to decline sequentially to 55.5% to 57.5%, or 2.1 points to 56.5% at midpoint. This is again due to continued FX headwinds, with approx. 2.6 points from unfavorable FX and overseas fab ramp in Kumamoto and Arizona.
Operating margin was guided to decline sequentially to 45.5% to 47.5%, or 3.1 points at midpoint. This would be the lowest level since Q2 2024.

EPS increased 67% YoY, up from 54%

TSMC delivered record profit in Q2, rising 61% YoY to NT$398.3 billion, or ~$12.8 billion. Adjusted EPS of $2.47 beat estimates for $2.31 and increased nearly 67% YoY, accelerating from recent growth in the 50% range. However, EPS growth is forecast to decelerate rather rapidly through Q4, with TSMC barely maintaining double-digit growth.

For 2025, adjusted EPS is expected to increase 35.2% YoY to $9.52, up from 31.5% growth two months ago.

Cash Flows dip yet Balance Sheet is Healthy

Cash flow margins dipped by a larger margin sequentially, and TSMC’s balance sheet remained healthy.

Operating cash flow was $16.2 billion for a 53.8% margin, down from a 74.5% margin in Q1 and a 56.1% margin in the year ago quarter.
Free cash flow was $6.5 billion for a 21.7% margin, down from a 35.1% margin in Q1 and a 25.5% margin in the year ago quarter.
Cash, equivalents and marketable securities totaled $90.4 billion, while debt totaled $32.3 billion.
Capex rose more than 51% YoY to $9.6 billion, slowing from a 74% pace in Q1. TSMC maintained its full year capex guide at $38 billion to $42 billion.
Inventories were $10.43 billion, up from $8.83 billion in Q1; however, the sequential increase looks to have been impacted by FX, as inventories in NT$ were up less than 4% QoQ.

Earnings call Q&A

While management offered little to no clarity on long-term AI growth or CoWoS capacity, they discussed long-term diversification of advanced node manufacturing to the US. Management also offered insights into advanced node capacity that signal Nvidia’s growth could remain strong come 2H.

Arizona Expansion

TSMC provided an update on its global expansion plans, which is important considering onshoring US manufacturing helps reduce geopolitical risk from China for the AI server supply chain. Management shared that they are expecting to bring 30% of their 2nm manufacturing to the US in Arizona.

As a result, TSMC is accelerating and expanding its presence in Arizona with its recent $165 billion investment plan, for six fabs, two advanced packaging fabs, and a major R&D facility to meet high multi-year demand from customers. The second fab in Arizona, utilizing 3nm tech, has finished construction with TSMC aiming to speed up volume production by several quarters. The third fab, offering 2nm and A16 advanced nodes, is under construction.

Management added that “despite the higher cost of overseas fabs, we will leverage our increasing size in Arizona and work on our operations to improve the cost structure,” to help minimize gross margin dilution impacts. This is important considering TSMC is forecasting increasing margin dilution from its growing overseas presence, widening from “2% to 3% every year in the early stages and widen to 3% to 4% in the later stages” over the next five years.

Nvidia H20 Impact

Morgan Stanley’s Charlie Chan questioned about Nvidia’s approval to resume shipping its H20 GPU to China, and if unlocking the Chinese market again would help TSMC raise its AI accelerator growth CAGR upwards from the mid-40% range.

C.C. Wei was rather tight-lipped about the potential impact, given that shipments have (likely) not yet resumed, saying that it is too early to provide an estimate on how this would impact growth. Wei explained that TSMC is not yet ready to increase its forecast, and “another quarter probably will be more appropriate to answer your question,” hinting that the AI accelerator CAGR may be updated in Q3.

However, it is expected that a majority of the H20s to be sold will be from existing inventory in the Taiwanese supply chain, meaning chips already built and revenue already booked. Therefore, it’s hard to see how TSMC could meaningfully increase its CAGR for the next four years based on just the H20.

In terms of China, however, the bigger opportunity here for TSMC may stem from Nvidia’s China-specific Blackwell B30 GPU, which is estimated to see shipments of up to 1.2 million units, or ~20% more than the total estimated H20 inventory. The B30 is expected to hit the market in Q4, following the H20’s resumption largely in Q3.

Advanced Node Capacity

Goldman Sachs’ Bruce Lu asked management about advanced node capacity, and supply-demand imbalances as more AI chips begin to shift to the 3nm node. While TSMC did not offer much beyond capacity being tight, one comment suggested Nvidia’s demand remains very strong.

C.C Wei explained that TSMC’s 5nm capacity is “very tight,” while 3nm capacity is even tighter and will continue to remain tight for a couple of years. Wei explained that TSMC can quickly retool advanced node fabs, such as 7nm to 5nm, 5nm to 3nm, etc, and keep utilization high to help meet demand. He would not commit to saying that demand would outpace supply, but that was implied given his comments of trying to “narrow the gap” between supply and demand.

However, one of the more important comments here was Wei stating that TSMC is using 7nm capacity to support 5nm demand, which provides another piece of evidence alongside surging HPC revenue that Nvidia’s Blackwell ramp is accelerating rapidly. Blackwell is built on TSMC’s N4P process, a subfamily of its 5nm node, offering higher performance, better power efficiency and higher transistor density.

Nvidia CEO Jensen Huang had stated that Nvidia was shipping nearly 1,000 racks per week to hyperscalers in May and expecting to ramp further, and this comment from TSMC supports lasting 5nm demand, likely from Nvidia given its high share of CoWoS (and manufacturing) capacity.

Conclusion

TSMC’s earnings provided more evidence that AI GPU demand remains strong, particularly for Nvidia, with HPC revenue rising to a fresh record with its largest QoQ increase, and commentary for 7nm capacity helping meet high 5nm demand. Supported by this robust AI and HPC demand, TSMC boosted its full-year guidance from mid-20% revenue growth to close to 30% growth, despite lingering concerns of tariff-related weakness come Q4.

Damien Robbins, Equity Analyst at I/O Fund contributed to this analysis.

Recommended Reading:

The I/O Fund’s Top 15 Stocks for Q3 2025

Posted on July 23, 2025June 30, 2026 by io-fund

This quarter, I’d like to try something new by providing Members with something more actionable than a 1-hour webinar on key trends. Trends are important to cover, yet I also realize one of my main roles is to provide our Members with stock picks. The stock picks below represent the trends that are in play, and thus, the list answers quite a bit of the questions around how we plan to position and why.

The analysis tops out at 42 pages and 16,300 words and it took about three weeks to write; there was no stone unturned to come up with this list.

As you know, we are a portfolio that is managed in real-time. This means that we are constantly evaluating how our content informs a portfolio. Ultimately, we believe taking time to produce the analysis below will make the I/O Fund portfolio stronger, and we hope it does the same for you.

A few disclaimers:

This is determined by lagging indicators and incorporates last quarter’s financials. There are many risks each of the stocks face below, and the analysis below is not a substitute for using technicals to guide an entry. For example, it’s very common for great tech stocks to selloff 40% or more about roughly every 12 to 24 months – each investor will need to determine how they plan to handle the tech sector’s inevitable volatility.

In addition, the broad market is currently at all-time highs; with this comes stretched valuations particularly from any stocks that are popular or well-known. We’ve strived to bring you an objective analysis which includes some stocks with admittedly stretched valuations that we hope to buy lower, as well as those that are lesser-known and sitting at lower valuations despite having a 10/10 thematic profile.

We have divided 15 stocks into three sections:

Quality AI for longer-term holds: four stocks ranked
AI Hardware plays (medium-term hold): six stocks ranked, two honorable mentions
AI software: six stocks ranked, two honorable mentions

Combined, we believe these stocks will represent about 60-65% of the I/O Fund portfolio. The remaining 35% to 40% will be allocated to crypto, energy and momentum stocks of any tech sector. These are general guidelines. If crypto were to lead in a meaningful way, we’d allocate more to this sector or if crypto tops, we will lower this allocation and rotate into quality AI, etcetera.

Please also note, the stocks are ranked for their respective categories. It would be nearly impossible to rank these perfectly as market forces are unpredictable. Most importantly, the true rank of stocks is found in our portfolio positioning.

The goal of this exercise is to more closely align our portfolio with trends in play and with fundamentals that are on fire. As you can see, this is one of the lengthiest reports I have ever compiled weighing in at over 16,000 words – but yet, there’s more. In addition to this fundamentals-driven report, Advanced Members have a lengthy incoming Quarterly Positions Report coming in from Knox where he matches the stocks listed below with technical setups.

Major themes:

Nvidia Blackwell: we’ve covered the importance of the upcoming generation of GPUs thoroughly beginning a year ago in the analysis “Blackwell and the $200B Data Center” plus “Here’s Why Nvidia Stock Will Reach $10 Trillion Market Cap.” We’ve then helped to clarify timing around Blackwell arriving with an analysis in February pointing to signals the premiere SKU was delayed and then further analysis in May stating we believe those signals have cleared and Blackwell is now ramping on the premium side.
Regardless of when the delay finally clears (August call or November call), I continue to believe that Nvidia and related suppliers are the best way to position into the second half of the year. We can see ample evidence that Blackwell is ramping, and what will catch the market off guard, is that Blackwell Ultra is preparing to ramp quickly after over about a 6-month to 9-month time frame. It isn’t confirmed yet when Blackwell Ultra will ship in volume but we do know that CoreWeave was first to market with a Blackwell Ultra deployment in early July. What's important here is the clock has started – while simultaneously Blackwell is ramping in volume.

You can think of this as a one-two punch that includes the supply chain in terms of Nvidia saying to the doubters: “you haven’t seen anything yet.” Blackwell’s strength against AMD and custom silicon lies in the demand for Nvidia’s NVL72 systems, which combine 72 GPUs using NVLink and NVSwitch to function like one massive accelerator. These systems are particularly attractive to hyperscalers looking for an edge in model training and high-throughput inference.

As of now, competitors like AMD and custom ASIC providers do not yet offer a comparable system that scales to this degree. Such capabilities may emerge around 2026–2027, but Nvidia currently holds a clear lead in large-scale, multi-GPU system integration. This is key to why 2025 belongs to Nvidia, with the second half backloadedwith the second half backloaded.
AI is diversifying: As we look further out into 2026, we’ve been preparing our Members for the inference market to cause capex to diversify away from Nvidia toward cheaper GPUs and custom silicon. Yet we also published on why it won’t matter for Nvidia in the long run. Briefly, there is plenty of money pouring into AI use cases for Nvidia to do quite well with a lower percentage of the AI accelerator market. The company has mid-90% of the market today, I believe Nvidia will end the decade in the mid-70% or low-80% of the market based on what we see in other markets such as gaming. Nvidia will make up for the lower market share with AI software and automotive, for example, which has been forecast to be a bigger market than AI hardware once the market matures.
Quality fundamentals but stretched valuations: Look for many popular stocks to struggle at these valuations to push to the next level. We are looking more broadly at stocks (listed below) that present lower valuations to help offset this risk. There are stocks we plan to aggressively buy while there are others we prefer to wait for a lower valuation. Regardless of valuation, we are listing our Top 15 with the understanding some have buy plans at lower levels while others are entering a buy zone very soon.
Delayed market reaction to tariffs: we continue to believe tariffs pose a risk to the market and although anything could cause incoming volatility for the tech sector (pick a headline), the fact remains that it’s a tall order for earnings to overcome margin pressures and potentially slowing demand. The first quarter had the benefit of a pull forward; the upcoming quarter will likely show some sectors reporting compressed margins.

Section 1: Quality AI Stocks (Long-term buy and hold)

There are four stocks in this section ranked from #1 to #4

1. Nvidia: The AI Leader on the Precipice of Round Two

Thematic: 10/10
Fundamentals: 10/10
Valuation: 5/10

Brief Overview:

You can view an interview on Fox where I discussed the puts and takes going into last quarter’s earnings report plus the new price target I/O Fund published here. The major takeaway is that Blackwell has enough ammo to push the stock into the mid-to-high $200s or a $6+ trillion market cap.

The second most important update on Nvidia this quarter is the commentary around Blackwell shipping in volume. According to our previous analysis: “On average, major hyperscalers are each deploying nearly 1,000 NVL72 racks or 72,000 Blackwell GPUs per week and are on track to further ramp output this quarter.” The rough math here implies hyperscalers are deploying $3 billion every week right now since each rack goes for $3 million. Furthermore, the run rate of this comment implies data center revenue will be above and beyond analyst consensus for Q2, Q3 and Q4 – thus, either analyst consensus comes up or these systems will become further supply constrained somewhere down the line and analysts are being conservative for now.”

Third, Jensen Huang is calling for exponential growth in inference. You will hear our firm discuss why this market opens up an opportunity for other players such as AMD and Broadcom, yet I want to make sure that comment is not lost in interpretation as its also quite bullish for Nvidia.

During the earnings call, Huang stated inference is reaching an inflection point, stating “we've reached an extraordinary milestone with AIs that are reasoning, are thinking, what people call inference time scaling. Of course, it created a whole new — we've entered an era where inference is going to be a significant part of the compute workload.”

He later also stated:

“Yeah, thanks. Thanks, Ben. I would say compared to the beginning of the year, compared to GTC timeframe, there are four positive surprises. The first positive surprise is the step function demand increase of reasoning AI, I think it is fairly clear now that AI is going through an exponential growth, and reasoning AI really busted through […] So, number one is inference reasoning and the exponential growth there, demand growth.”

Overall Revenue Growth:

Last quarter was mired by the loss of China revenue, yet the company still managed to report a slight revenue beat in Q1, reporting 69.2% YoY growth to $44.06 billion in revenue, just ahead of the $43.25 billion consensus.

AI Segment Revenue Growth:

Nvidia reported 73.3% growth in data center revenue to $39.11 billion in Q1, marginally higher than analyst expectations from Visible Alpha of $39.08 billion. This marked the end of Nvidia’s seven-quarter streak of $1 billion-plus beats in the segment – based on the Visible Alpha estimate, Nvidia beat by just $33 million, its lowest in the past nine quarters. This makes sense considering they are in-between GPU generations.

Compute revenue rose 76% YoY but just 5% QoQ to $34.16 billion, impacted by the H20 ban, while Networking revenue rebounded swiftly, rising 56% YoY and 65% QoQ to $4.96 billion. Nvidia said Networking’s performance was “driven by the growth of NVLink compute fabric in our GB200 systems and continued adoption of Ethernet for AI solutions at cloud service providers and consumer internet companies.”

Earnings:

Nvidia reported a slight EPS beat despite the margin contractions, with adjusted EPS of $0.81 coming in ahead of the $0.75 estimate. GAAP EPS of $0.76 missed estimates for $0.81.

Adjusted EPS growth slowed quite dramatically, decelerating more than 38 points sequentially, in part due to the H20 ban; Nvidia noted that excluding the ban, adjusted EPS would be $0.96. This would represent YoY growth of 57.4% versus the 32.8% reported.

Looking ahead, adjusted EPS growth is expected to rebound and remain in the low to mid-40% range as margins recover. However, given that Q1’s EPS excluding the ban showed growth in the high-50% range, estimates may move higher as Q2’s margin outlook shows almost no persisting impact.

Margins:

GAAP gross margin was 60.5% and adjusted gross margin was 61%, around 10 points below management’s initial guidance for 70.6% and 71% due to the $4.54 billion charge related to the H20 ban. For Q2, management guided for 71.8% GAAP gross margins and 72% adjusted gross margins, a rebound of approx. 11 points sequentially.
GAAP operating margin was 49.1%, well below guidance for 58.5% and a sequential contraction of 12 points. Adjusted gross margin was 52.8%, nearly 10 points below the guide for 62.6% and a sequential contraction of more than 12 points.
For Q2, management’s guidance implies operating margins will rebound with gross margins, projecting approximately a 10 point sequential expansion to a 59.1% GAAP and 63.1% adjusted operating margin.
GAAP net margin was 42.6%, while adjusted net margin was 45.2%. The broad-based margin recovery in Q2 is expected to mostly transfer through to the bottom line, with management guiding for a 7.6 point recovery to a 50.2% GAAP net margin.

Cash:

Cash flows were surprisingly strong as Nvidia’s cash flow margins expanded approximately 20 points sequentially, while it added more than $10 billion in cash to its balance sheet.

Operating cash flow was $27.41 billion, up nearly 79% YoY on higher revenue. OCF margin was 62.2%, up 20 points QoQ and more than 3 points YoY.

Free cash flow was $26.14 billion, up 75% YoY. FCF margin was 59.3%, up nearly 20 points QoQ and just 2 points YoY.

Valuation:

Nvidia trades at a 20 forward PS with its minimum sales ratio at 10 and maximum around 30. Therefore, the stock is in the mid-range. The stock trades at a 38 forward PE ratio, which is also right in the middle of its trading history, seeing a minimum of 20 and a maximum of 48 in recent years.

Where the edge in Nvidia’s valuation lies is the sudden release of Blackwell Ultra following Blackwell. This should create a new, upward trajectory in revenue growth (if we assume supply chains cooperate) since there will be a historic, back-to-back release in two monumental GPU generations. To put it another way, Blackwell’s delay caused a year of flat price action but given Nvidia continued to develop its next generation Blackwell Ultra during that delay, probability favors us seeing a $6 trillion market cap sometime next year.

Notable Risks:

Perhaps a $6 trillion market cap sounds fancy yet there will be higher returns in choice Nvidia suppliers as the stock is already at $4 trillion. The risks to Nvidia are low, yet there is opportunity with a quality stock especially when there are many beneficiaries a bit further down the supply chain that will see hypergrowth for Blackwell and Blackwell Ultra.

2. TSM: The Stock with More Pricing Power than Even Nvidia

I try to not use the word “moat” too loosely, yet TSM is deserving of this recognition for its deep IP and market lead. TSMC continues to deepen its moat with advanced nodes, such as N2 and A16. The company already powers tens of trillions in market cap on the stock market when you consider Apple, Nvidia, Broadcom, Amazon, AMD and Google are customers of TSMC. Essentially, all mega cap stocks have an AI strategy spanning merchant GPUs and custom silicon, and of course, software – yet the common denominator to these strategies is they all funnel into TSMC.

The most advanced node shipping today is the 3nm, offering 15% better performance than the 5nm process when power level and transistors are equal. The die sizes are an estimated 42% smaller than the 5nm and TSMC also states the 3nm process can lower power consumption by as much as 30%.

Power efficiency is a major advantage, helping to deepen TSMC’s moat. Samsung was first to introduce 3nm process chips in 2022 yet has not been as competitive on yield and power efficiency at a roughly 10% to 20% difference compared to TSMC. The moat is visibly seen in TSM’s pricing power with the dominant foundry charging 25% more for its 3nm process compared to its 5nm process, and customers are willing to forego Samsung to pay the higher pricing.

Last year, companies such as Apple, Nvidia, AMD and Intel committed to working with TSMC for its 3nm process, and eventually Google and Qualcomm left Samsung “after careful consideration” to also secure a partnership with TSMC.

This was an important moment for TSMC to complete its near-monopoly in advanced nodes as Google had been outsourcing its Tensor processors to Samsung’s foundry for four generations, before moving to TSMC for the fifth generation. Qualcomm also switched to TSMC from Samsung for the Snapdragon 8 Gen 4 series.

To attract these large customers with different end markets, TSMC offers a few 3nm processes, such as the N3E, N3P and N3X. This allows a company like Apple to customize the 3nm chips differently than AI chips for hyperscalers. N3E is the baseline for IP design with 18% increased performance and 34% power reduction, N3P has higher performance and lower power consumption, whereas the N3X will offer high-performance computing very high performance but with higher power leakage.

To illustrate the near monopoly that TSMC has over other foundries, consider that its market share stands at 67.1%, up 2.4% QoQ in Q4. Meanwhile, second-place Samsung was at 8.1% down from 9.1% for a lead of 59 points.

When comparing revenue, TSMC reported $26.85 billion in Q4 for a 14.1% increase compared to Samsung’s $3.26 billion, which declined 100 basis points to 8.1%. The most recent quarter, TSMC furthered the lead with $30.7 billion in revenue.

Overall Revenue Growth:

TSMC offers monthly revenue reports, providing a high level of visibility into the chipmaker’s growth. For example, May revenue rose 39.6% YoY to NT$320.52 billion (~US$10.7 billion), while June revenue rose 26.9% YoY to NT$263.71 billion (~US$9.0 billion). For the first half of the year, revenue rose nearly 40% YoY to NT$1,773.05 billion (~US$55.6 billion).

Q2 revenue also outperformed the company’s guidance of $28.4 – $29.2 billion, rising 44.4% YoY to $30.07 billion, driven by AI and HPC products with some FX tailwinds. Q3 is expected to remain strong with 37.8% YoY growth to $32.4 billion at midpoint.

Regardless of which way you dice it, TSMC is guiding for above industry growth, updating its 2025 guidance in the most recent quarter from mid-20% YoY growth to close to 30% YoY growth.

Of this, AI accelerator revenue is expected to double in 2025 and management also forecasts AI to grow at a mid-40% CAGR for five years from 2024: “Based on our planning framework, we are confident that our revenue growth from AI accelerators will approach a mid-40s percentage CAGR for the next five years period starting from 2024.”

AI Segment Revenue Growth:

TSMC continues to ride AI accelerator tailwinds, evident in its rising HPC revenue and mix. HPC revenue continued to accelerate in Q2, now reaching $18 billion, marking its largest QoQ increase of nearly $3 billion. HPC accounted for 60% of TSMC’s revenue, expanding slightly from 59% of revenue last quarter. Management stated that they are continuing to observe robust AI-related demand from customers with no change in behavior, despite lingering tariff-related concerns.

In the latest quarter, advanced nodes below 7nm drove 74% of wafer revenue with 3nm contributing 24% of revenue and 5nm representing 36% of revenue. Nvidia is not on the 3nm process yet for its Blackwell shipments, thus 5nm is outsized in terms of its market share.

Earnings:

TSMC delivered record profit in Q2, rising 61% YoY to NT$398.3 billion, or ~$12.8 billion. Adjusted EPS of $2.47 beat estimates for $2.31 and increased nearly 67% YoY, accelerating from recent growth in the 50% range. This EPS growth also reflects TSMC’s operating leverage, outpacing revenue growth by 23 points.

For Q3, EPS growth is expected to increase 27% to $2.46, before decelerating rather sharply to barely in the double-digits by Q4 as it begins to lap these more difficult 50% growth comps.

For 2025, adjusted EPS is expected to increase 35.2% YoY to $9.52, up from 31.5% growth two months ago. Growth is forecast to decelerate to 14.4% in FY26 to $10.89.

Margins:

Similar to Nvidia and Broadcom, TSM has excellent margins:

Gross margin was 58.6% in Q2, at the high end of the guided range for 57% to 59%. Gross margin declined sequentially from 58.8%, with a 2.2 point headwind from FX and a 1 point headwind ramping overseas fabs offset by higher capacity utilization.
Operating margin was 49.6%, increasing 1.1 points sequentially from operating leverage, and above guidance for 47% to 49%. Operating margin was up more than 7 points YoY.
Net margin was 42.7%, down slightly from 43.1% in the prior quarter but up nearly 6 points YoY.

Over the next five years, management sees the dilutive impact from ramping its overseas fabs widening, projecting it to start at 2-3% each year in the early ramp stages before widening to 3-4% each year. Despite this, TSMC remains confident in its ability to keep long-term gross margins at 53% or higher.

Cash:

Free cash flow was $6.5 billion in Q2 for a 21.7% margin, down from a 35.1% margin in Q1 and a 25.5% margin in the year ago quarter. This is at the lower end of the typical range for TSMC, which tends to track between 20% to 30%.

Valuation:

TSM is trading at the high end of the range on the bottom line at 24 forward PE ratio, with its highest being 30 over the past year before it saw a sharp adjustment to 15 forward PE Ratio, which marked a bottom. Forward PE Ratio of 20 is mid-range and more comfortable for this stock.

On the top line, TSM is trading at 10 forward PS with 12 marking a top over the past year and 6 marking a bottom. This leaves very limited upside.

Risks:

Valuation tops the list as the primary risk given the stock is rather insulated from competition, and is also insulated from any tariff drama as it has the best pricing power in the industry. Onshoring its fabs with continued stimulus helps to create a more durable stock.

3. Broadcom: Quietly Reached $1 Trillion, Will Displace the FAANGs

Thematic: 10/10
Fundamentals: 10/10
Valuation: 1/10

Similar to Nvidia and TSM, Broadcom will likely remain on the Top 15 list for years to come and will at times outrank Nvidia.

Broadcom stock joined Nvidia, Alphabet and Microsoft in calling out surging AI inference demand, noting that this rapid growth could drive increased demand for custom silicon in the second half of 2026, and with it, higher AI revenue.

Despite an in-line print and guide, Broadcom’s AI revenue is tracking above Street estimates for next year towards the $30 billion mark, up nearly 150% in two years, with growing tailwinds from inference and networking as clusters increase in size. AI revenue growth is also tracking Broadcom’s addressable market forecast of a 60% CAGR.

Broadcom is cementing itself as the clear second in AI with key ingredients for success as inference demand rises. However, its premium valuation to Nvidia looks to be pricing in above-expected AI revenue growth into 2027, likely closer to a 70%+ CAGR, as there exists a $160 billion gap in AI-driven revenue between the two.

HSBC estimates that Broadcom’s ASIC revenue could rise as much as 128% YoY next year to $28.3 billion, fueled by Google’s TPU ramp driving a 92% YoY increase in ASIC ASPs. – xAI Beth_Kindig xAI Beth_Kindig

Overall Revenue Growth:

Broadcom reported $15 billion in revenue versus $14.99 billion expected, up 20%. Management expects about $15.8 billion in third-quarter revenue, versus $15.7 billion.

AI’s strength is masking persisting softness in non-AI revenue, which could continue to be pressured due to Broadcom’s high consumer exposure. Broadcom noted that non-AI revenue “is close to the bottom” but it “has been relatively slow to recover” with revenue down (5%) YoY to $4 billion in Q2.

AI Segment Growth:

Earnings:

Broadcom reported $1.58 adjusted versus $1.56 expected.

Margins:

Broadcom has excellent margins especially given its scale as primarily a hardware company:

GAAP Gross Margin: 68% and adjusted gross margin of 79.4%

GAAP Operating Margin of 38.8% and Adjusted operating margin of 65.3%

Net Margin of 33% and Adjusted net margin of 51.9% for adjusted net profit of $7.8 billion.

Cash:

Broadcom has seen expanding cash flows with operating cash flow margin of 43.7% up from 36.7% last year. Free cash flow of 42.7% compares to 36.7% last year for free cash flow of $6.4 billion. The company had $9.48  billion at quarter-end

Valuation:

Broadcom has a cringe-worthy valuation at 42 forward PE ratio and 21 forward PS. You can argue it’s worth as much as Nvidia given its AI growth, yet there is a whopping $160B delta between Nvidia and Broadcom’s AI sales. Broadcom is traditionally a 6 forward PS company and a 2 current PS ratio. Keep in mind, Broadcom supplied Apple during the mobile boom and outperformed most FAANGs, therefore, even with being in the center of a microtrend the stock is overvalued according to most standards.

Notable Risks:

The valuation on Broadcom stands out as one of the most egregious on our list, second only to Palantir and perhaps tied with Cloudflare. Some investors buy stretched stocks successfully for a period of time and believe they’ve outsmarted the market; however, I do not see these valuations sustaining, and in fact, these three point more toward a bubble of sorts.

You can read more in the analysis: “This AI Stock is Set to Surge from Inference Demand” on the free side including a follow-up with information on the premium side.

4. AMD: The Dark Horse in our Stable; Patience is a Virtue

Last month, AMD introduced its Instinct MI350 series GPUs, including MI355X with up to 4X performance over the previous MI300X generation and up to 40% more tokens per dollar compared to Nvidia’s B200 accelerators. The company also previewed its Helios rack-scale server architecture featuring the MI400s for 2026 deployments.

According to Tom's Hardware AMD is claiming the eight-GPU MI355X system is 1.3X faster than Nvidia’s DGX GB200s systems with Llama 3.1 and up to 1.2X faster than the B200 HGX systems in inference for DeepSeek R1 with equivalent performance as Llama 3.1 when tested at FP4.

Perhaps what matters most to investors is what the GPUs will cost. The team has been digging around and found the following this week:

HSBC last week upgraded AMD to Buy and doubled its price target to $200, saying that it expects the MI355 GPUs [and MI400s] to command a $25K ASP, up materially from prior assumptions for $15K, potentially driving upside to FY26 AI revenue – xAI Beth_Kindig xAI Beth_Kindig

Overall Revenue Growth:

AMD reported a double beat in Q1 with revenue of 36% and data center growth of 57%, with the beat filtering down to the bottom line with EPS growth of 55% — ahead of revenue.

AMD reported Q1 revenue of $7.44 billion, solidly ahead of the $7.12 billion estimate and above the upper range of its guidance for $7.1 billion, +/- $300 million. Revenue growth accelerated to 35.9% YoY, led by data center and client, although as of now, this is expected to be the peak growth quarter for the year.

On one hand, it is quite impressive AMD can overcome the impact from China last quarter and meet consensus for next quarter. On the other hand, analysts have been lowering estimates as AMD was supposed to see revenue of $7.77 billion for growth of 33% as of last October rather than the 26.7% in the current quarter.

Arguably, the news that Nvidia can resume sales of H20s may be bigger news from AMD if we assume the $1.5B impact is removed from AI revenue, as it's 30% of AI revenue as it stands for this company versus less than 10% for Nvidia (at $15B versus somewhere around $150B for rest of year in AI revenue).

However, if we look at the facts on the table, AMD is a weaker stock overall than Nvidia, TSM and Broadcom as it stands today. It is my speculation this changes and AMD shows its full potential in 2026-2027 aligned with the inference market and the shift in priority where Big Tech becomes more cost conscious.

AI Segment Growth:

Data Center revenue grew 57% YoY but declined (5%) QoQ to $3.67 billion, driven by sales of EPYC CPUs and Instinct GPUs, accounting for over 49% of AMD’s revenue in the quarter.

While growth decelerated from 69% in Q4, it’s coming against a much tougher comp at 80% YoY whereas Q4 of last year offered a lower comp of 38% in Q4 2023.

Regarding GPUs, management stated their AI revenue increased by a “significant double-digit percentage year-over-year.” The MI325X is shipping in volume while the next-gen Instinct MI350-series chips are on track for “accelerated production by mid-2025.”

We discussed last quarter that AMD was pushing up their delivery on the MI350s to mid-year for relative competitiveness. For Q2, data center will decline due to the MI308 revenue being excluded.

When asked about future quarters, the CEO Lisa Su stated the DC segment would resume growth after Q2: “in Q2, it's not going to grow year-over-year just given what we've said about the $700 million coming out of Q2 and how we had previously talked about the evolution. But we do believe that we'll grow year-over-year going forward, in Q3 and Q4 certainly, for us to do the full year with strong double-digit growth.”

To put it plainly, on the AI accelerator front, this will be the first time that AMD will overlap Nvidia in terms of benchmarks on GPUs. Please do note, the amount of time that AMD’s current generation of GPUs and Nvidia’s GPUs overlap will be brief – and will only be at the single GPU and 8-GPU system level. AMD was originally expected to ship the MI350s at the end of this year yet are moving the shipments up – which fits with AMD’s tradition of underpromising and overdelivering.

However, the accomplishment is noteworthy as it’s setting the tone as the inference market begins to ramp. In other words, AMD ceded the training market to Nvidia – but I do not expect that to be the case with the inference market.

When Blackwell Ultra ships, the B300s will offer FP4 TFLOP/s that is 1.3X faster than AMD’s current MI350X and MI355X. With that said, because AMD has prioritized competing on memory — its bandwidth and capacity is expected to be on par with Blackwell Ultra.

The market is forward-looking, which means investors should be too. AMD is closing the gap on single GPUs and 8-GPU systems, yet the MI400s will mark a pivotal moment as AMD will attempt to compete on rack-scale systems with Helios, its 72-GPU systems. If things go as planned, AMD will be competitive with Nvidia on GPU, memory and interconnect performance — while potentially taking the lead on memory capacity and bandwidth.

Earnings:

AMD reported adjusted EPS of $0.96 in Q1, slightly ahead of estimates for $0.93. This represented YoY growth of 54.8%, accelerating from nearly 42% growth last quarter. Similar to revenue, Q1 is currently expected to be peak growth for EPS, with Q2 estimated to record 27.8% growth before slowing to the low 20% level by Q4.

However, management commented that EPS growth is expected to grow much faster than revenue in Q2: “Looking at Q2, at the middle point of our guidance, revenue will be increasing 27%, and we do expect the earnings per share growing much faster than the top line revenue growth.”

Margins:

AMD’s margins are not nearly as strong as Nvidia or Broadcom’s, and this is ultimately reflected in the valuation. However, should the HSBC analyst quoted above be correct (and general consensus that AMD has some kind of pricing power), the margins will improve over time as GPU sales ramp at higher ASPs.

Q1 GAAP gross margin was 50%, up 3 points YoY, while adjusted gross margin was 54%, up 2 points YoY.
GAAP operating margin was 11%, a strong expansion of 10 points YoY, and adjusted operating margin was 24%, up 3 points YoY.
GAAP net margin as 9%, up 7 points YoY, and adjusted net margin was 21%, up 2 points YoY.

Cash:

Operating cash flow was $939 million for a 13% margin, expanding from a 10% margin in the year ago quarter.

Free cash flow was $727 million for a 10% margin, expanding from a 7% margin a year ago.

Expect these to improve once the AI story ramps.

Valuation:

AMD is trading mid-range at forward PS of 7. The stock can trade as low as 4 or high as 12. The stock is trading at a 40 forward PE Ratio, which is also mid-range given it’s traded as low as 25 and as high as 55 in the past year.

Notable Risks:

The risk to AMD is primarily in Q2’s data center growth decline, and how quickly can the company ramp its MI355s and subsequent MI400s while in the midst of Nvidia’s large shadow – will we see a solid surprise arrive in Q3, Q4 or even into next year? My best guess is the most meaningful AMD moment is not likely to occur during Blackwell’s NVL72s release – I think 2025 belongs to Nvidia and somewhere between 2026-2027 we switch it up.

My current prediction is that AMD will offer higher stock returns than Nvidia by 2028, which you can read more about here in the analysis: AMD vs Nvidia: The AI Stock that Could Win by 2028.

Section 2: AI Hardware Plays (Medium-term hold)

This section has six stocks ranked #1 to #6 with two honorable mentions

1. Astera Labs: AI Networking Pureplay Serving Two Enormous TAMs (total addressable markets)

Thematic: 10/10
Fundamentals: 10/10
Valuation: 5/10

Astera Labs reported an impressive beat and raise in Q1, with GAAP margins strengthening as revenue continues to grow at a triple-digit rate. On top of this impressive beat, the growth story for Astera Labs is only beginning. The commentary regarding their product diversification and higher dollar content going into the second half of the year was quite clear as to the growing opportunity this company is poised to capture.

Primarily, Astera offers unique positioning that allows them to capture both the merchant GPU market and custom silicon market across its three products lines Astera, Taurus and Scorpio. This widens the TAM and allows for steady revenue growth despite hiccups or delays from a single AI system (which we’ve seen plenty of disruption recently across those with high customer concentration with Nvidia).

In addition to being a strong custom silicon vendor for hyperscalers, Astera will participate in Blackwell once it (finally) ships in volume as the company offers PCIe scale-out and Ethernet scale-up. Their new products Scorpio P-Series and Scorpio X-Series are fabric switches that are particularly well-suited for the immense demand that is expected for customization of racks as architectures scale-up in the second half of the year and beyond.

Notably, Aries PCIe retimers and Taurus Ethernet smart cable modules are driving the revenue today with the Scorpio P-Series beginning to ramp. However, there are many catalysts on the horizon for Astera which adds to the trifecta of a strong growth story:

Serving both ASICs and GPUs greatly increases TAM and diversifies revenue; rare in the AI systems ecosystem
Preparing to serve the scale-out demand with increasing higher dollar content; specifically on Scorpio but also on Aries
Offering strong cross-sell opportunities as it aims to be the first to solve unique challenges for both GPU and custom silicon utilization – and is solving these issues in a way that avoids vendor lock-in for the large hyperscalers who want a mix of both custom silicon and merchant GPUs (Nvidia or AMD).

Astera Labs Fundamentals Update:

Overall Revenue Growth:

Astera Labs reported an impressive 144.3% YoY revenue growth in Q1 to $159.4 million, topping analyst estimates for $151.5 million in the quarter.

For Q2, Astera delivered a solid raise at $170 to $175 million, more than 7% ahead of the $160 million estimate. This points to YoY growth of 124.5% at midpoint, ahead of estimates for just 108% YoY. What’s impressive about this ramp is that Astera is guiding to deliver this 125% growth in Q2 against its 619% YoY comp (against a small base), for its seventh-straight triple-digit growth quarter.

Astera has seen revenue growth decelerate over the past few quarters, with growth expected to continue decelerating as Astera laps its rapid ramp quarters. What’s impressive about this ramp is that Astera is guiding to deliver this 125% growth in Q2 against its 619% YoY comp (against a small base), for its seventh-straight triple-digit growth quarter.

For the full-year, Astera did not provide a guide, though estimates heading into Q1’s report were pointing to 70.4% YoY growth to $675.2 million in revenue. However, given that Q1 and Q2 have combined for a $20 million beat compared to current estimates, it’s likely that full-year revenue estimates will likely move closer to (or above) $700 million in the coming days. This would correspond to YoY growth of nearly 77%.

Key AI Segment:

The Scorpio P-Series is shipping this quarter and are qualified for Nvidia systems, yet the X-Series will ship in H2 with a bigger opportunity for custom silicon clusters. The Scorpio P-Series is a small chip that connects the CPU, GPU, NIC and NVMe storage. Rather than building a large switch, the company built a smaller device that is more efficient for high-speed signals to help feed GPUs with data. The fewer ports and smaller switch decrease complexity in a bid to compete against Broadcom with twice the lane count.

Inventories rose 18.2% QoQ to $51.1 million, likely driven by the ramp of Astera’s Aries 6 and Scorpio P-series products.
Accounts receivable surged 100.5% QoQ to $69.8 million, driven by Astera’s largest customers.
Astera’s receivable balance from its top customer in the quarter rose 363% QoQ to $20.9 million, while its balances from its second and third largest customers rose 75% and 90% QoQ to $14.7 million each.
Days sales outstanding also increased from 20-ish days in the past to 40 days this quarter. This is likely foreshadowing Astera is preparing for larger shipments in the next 1-2 quarters

Latest report can be found here: Astera Labs: Product Differentiation is Set to Soar in H2 and Beyond

Earnings:

Astera delivered an impressive 350% beat to GAAP EPS estimates in Q1, driven by its operating margin expansion, while forecasting EPS above estimates for Q2.

Adjusted EPS of $0.33 beat estimates by $0.05, representing YoY growth of 230%.

GAAP EPS of $0.18 beat estimates by $0.14, improving from $0.14 in Q4 and marking its second straight quarter of GAAP profitability on the bottom line.

For Q3, Astera guided for adjusted EPS between $0.32 and $0.33, approximately flat QoQ but up 150% YoY at midpoint.

Margins:

Gross Margin = 75%
GAAP operating margin = 7%
Net margin = 20%

Astera is guiding for margins to remain strong in Q2, with GAAP operating margin expanding. Gross margin was guided at 74% once again, while GAAP operating margin is forecast at 7.9%, up 0.8 points sequentially. Adjusted operating margin is forecast to contract 2.6 points QoQ to 31.1%.

Cash:

Operating cash flow was $10.5 million for a 6.6% margin, expanding slightly from a 5.6% margin in the year ago quarter.

Free cash flow was $6.0 million, for a 3.7% margin, improving from a 0.3% margin in the year ago quarter.

Valuation:

Astera’s valuation is trading mid-range of its historic trends at 21 forward PS and could trade as low as 40% lower from here or could have up to a 200% upward move from the current valuation. There is no guidance from valuation on where the stock will go next whereas others are more visibly overstretched.

Notable Risks:

Coming up on tough comps, high SBC weighs on operating margin but gross margin is one of the highest in AI semis. Cash has a weak margin, but scale will likely resolve any cash flow margin issues

2. Credo: AEC Networking Tailwinds

Thematic: 10/10
Fundamentals: 10/10
Valuation: 5/10

Credo continues to report outstanding revenue growth, up 180% YoY in Q4 and guided to accelerate further in Q1 as management touted growing traction with hyperscalers, new design wins in qualification and strong customer forecasts driving sustained AEC growth.

GAAP margins have expanded significantly down the line with operating margin quickly approaching 20% as signs of operating leverage emerge. Cash flow margins were robust in Q4 on strong collections, while inventories surged over the past two quarters, indicating that Credo’s hypergrowth phase will likely continue for a few quarters.

Management hinted that a new DSP deal with a hyperscaler represents its largest revenue opportunity to date, with two new hyperscaler customers ramping up in FY26. Backed by these arising revenue streams, Credo guided for revenue growth of 85%+ next year, or over $800 million.

Latest report can be found here: Credo Reports 180% YoY Growth and 20% GAAP Operating Margin.

Overall Revenue Growth:

Note: Upcoming earnings is Q1 Fiscal Year 2026

Credo reported 179.7% YoY and 25.9% QoQ growth to $170.0 million in revenue in Q4, beating the consensus estimate for $159.6 million. Revenue growth has sharply accelerated throughout the fiscal year, up from the 60% to 70% level in 1H to high triple digits in 2H.

For Q1, Credo guided to $185 million to $195 million in revenue, pointing to a nearly 40 point sequential acceleration to 218% YoY growth at midpoint. This was also 17% above consensus estimates for $162.4 million heading into the report. Revenue growth estimates have moved sharply higher since February. Q1’s growth estimate just four months ago was 133.4%, and is now nearly 85 points higher, while Q2’s growth estimate has risen 74 points from 100.9%.

For fiscal 2025, Credo reported a 122 point acceleration to 126.3% YoY growth, with revenue of $436.8 million. For fiscal 2026, Credo guided for revenue to exceed $800 million, for growth in excess of 85% YoY, while analysts are now expecting $804.1 million.

Key AI Segment:

Credo reported a significant 80-point sequential acceleration in product revenue growth to 303.3% YoY in Q4, with revenue of $164.5 million. Credo said AEC products are gaining traction in rack-to-rack distances up to 7 meters, with xAI being the most successful customer at that distance with a second customer ramping this year.

For optics, Credo noted that it reached its revenue targets and ended FY on strong momentum with an expanding customer base. As previously mentioned, Credo is targeting 100%+ optics revenue growth in FY26. Moving forward, Credo expects to diversify its customer base, eyeing up to five >10% customers in FY26, up from three in FY25. Credo’s largest customer, rumored to be Microsoft, accounted for 61% of revenue in Q4.

Earnings:

Credo’s fiscal 2025 adjusted EPS of $0.70 increased from $0.08 in the prior year. Credo generated the bulk of this EPS in H2 as revenue and margins surged.

Adjusted EPS of $0.35 in Q4 beat estimates by 29.6%, representing growth of 400% YoY. Growth is forecast to accelerate to 782% in Q1 to $0.35 on a low comp, before slowing to 17% YoY by Q4 FY26 against a much tougher comp.

Margins:

Credo is GAAP Profitable.

Gross Margin = 60%
GAAP OM = 8.5%
GAAP Net Margin = 12%

Cash:

The company has expanding cash flows.

OCF margin was 34% in the quarter, compared to 3.1% last quarter and 6.8% a year ago. Operating cash flow was $57.8 million up from $53 million QoQ.
Free cash flow was $54.2 million in Q4, for a 31.9% margin.
For FY25, free cash flow was $29 million, for a 6.6% margin, down from an 8.9% margin last year on higher capex.

Valuation:

Credo’s valuation is trading mid-range of its historic trends at 21 forward PS and could trade as low as 40% lower from here or has another 50% upward move. There is no guidance from valuation on where the stock will go next whereas others are more visibly overstretched.

Notable Risks:

Coming up on tough comps, 75% exposure to Hong Kong, copper recently undercame new tariff laws and where Credo sources copper is unlikely to be of public record. GAAP OM could be better but gross margin is impressive for AI semi

3. Supermicro: Key Nvidia Supplier Sitting in Plain Sight

Thematic: 10/10
Fundamentals: 4/10
Valuation: 10/10

Super Micro, also known as Supermicro, is sandwiched in the AI trend between hyperscalers and major chip design companies. The company is a server maker that started off by making motherboards and other components before it began making complete systems. The company is unique in that it sits between being an equipment manufacturer (Dell, HP) and being a design manufacturer (Foxconn).

To give you an idea as to the company’s sudden ascent off the Hopper generation of GPUs from Nvidia, consider that SMCI had revenue of $2.5B in 2021 and reported $22 billion in the fiscal year ending in June – or about 9X in four years. Given AI servers are increasing in complexity, and will require thermal management including direct liquid cooling, this is a baseline of what SMCI will be capable of over the next few years. There may not be the sudden 9X trajectory we saw off small numbers, but there will likely be ample growth.

The company is not without risks. There was a high-profile accounting issues recently, and Supermicro also struggles with cash (potentially diluting shareholders down the line) and has slim operating margins.

I’m calling this one “sitting in plain sight” because its valuation is low relative to the opportunity. It's also apparent the market has overlooked not only Supermicro but is overlooking Nvidia’s Blackwell since it took much longer to arrive than originally anticipated.

Overall Revenue Growth:

Fiscal Q3 net sales were $4.6 billion, up ~19% year over year. However, this was 19% lower than the prior quarter and below management’s forecast due to delayed customer commitments (some clients postponed orders while awaiting new AI platforms)

AI Segment Revenue Growth:

AI-focused products drove the majority of sales. Management noted that AI GPU platforms accounted for over 70% of Q3 revenue. Supermicro achieved volume shipments of new AI server platforms.

Earnings:

Super Micro's EPS for the most recent quarter (Q3 FY2025) was $0.31. This figure represents the non-GAAP diluted net income per share. The company also reported a GAAP EPS of $0.17

Margins:

Profitability declined sharply. Gross margin fell to ~9.6% (versus ~15.5% in Q3 last year). Pressured by higher inventory reserves and lower volumes

Gross Margin (GAAP): 9.6%, down from 11.8% in the prior quarter and 15.5% year-over-year
GAAP operating expenses were $293 million, generating GAAP operating income of approximately $147 million (net income before taxes and interest), which equates to roughly 3.2% operating margin on $4.60 billion revenue
Net income of $109 million on $4.60 billion in sales yields a 2.4% net margin

Cash:

The company generated $627 million in operating cash flow during the quarter. It ended Q3 with $2.54 billion in cash (against $2.49 billion in debt), yielding a slight net cash position of about $44 million

Valuation:

Valuation is what makes this stock attractive. I believe the last earnings report was a “red herring” of sorts, meaning it does not represent the bull story, which is the incoming shipments from Blackwell. This means the fundamentals were depressed last quarter, further depressing the valuation.

Trading at 1 fwd PS is worth the risk, in my opinion. This stock should always have a trailing stop due to weak margins and weak cash (overall weak FA profile). However, the growth story should also not be ignored. Look for this stock to comfortably go to 2-3 fwd PS on the upper end, and as low as 0.5 fwd PS which would be a layered buy in addition to 1 fwd PS. Overall, I expect fundamentals and valuation to resume Hopper-generation status sometime in the next 6 months – which means max’ing out between 2-3 fwd PS and GAAP operating margins that are in the 10%+ range up from 3% operating margin now.

Notable Risks:

Supermicro has very poor fundamentals as it must raise cash to scale. Being a commoditized AI server/hardware company, the margins are slim to none. It’s not clear if domesticating supply chains will help SMCI’s margins (it could). SMCI offered a red herring type earnings report as the company’s results got slammed by previous generation GPUs (Hopper, Chinese servers) yet will likely do quite well from incoming Blackwell.

Note: there is no recent analysis on Supermicro as we are looking to add this stock to our portfolio after taking a pause for about a year on the stock. You can find our previous thesis from 2023 on Supermicro here.on Supermicro here.

4. Dell: Strong Initial Sales from Blackwell with 612% Growth in Backlog

Thematic: 8/10
Fundamentals: 7/10
Valuation: 5/10

There are a few key catalysts to keep an eye on for Dell’s growth story. First off, will Dell move from primarily enterprise servers to also supply hyperscalers with AI servers? The current margin profile suggests this may already be happening as Tier 1 hyperscalers demand lower margins than enterprise servers. Meta, xAI and Coreweave are confirmed customers; the question is if the Big 3 follow.

AI factories are a major growth story for Dell – defined as complete systems that bundle PowerEdge AI servers, high-performance storage, intelligent networking, and integrated software/services. There is higher dollar content and higher margins on the storage and networking side for Dell as NVIDIA’s Blackwell GPUs and Dell’s cooling and integration expertise are combined to offer on-site (on-premise) AI servers.

Dell shows a lower thematic rating than peer Supermicro because it has a large Client segment. Dell has a higher rating on fundamentals due to its strong cash flows (a pain point for SMCI) and for its reliable management team, who are experienced at running a profitable company at scale.

Overall Revenue:

Dell reported $23.38 billion in revenue in Q1, a slight <1% beat to estimates as all of its core businesses grew in the quarter.

Revenue growth decelerated to 5.1% YoY in the quarter with Dell forecasting a sharp acceleration in Q2 as it is now rapidly ramping AI server shipments after orders surged in Q1.

For Q2, Dell guided $28.5 to $29.5 billion in revenue, or 15.9% YoY growth at the $29 billion midpoint, which marks a nearly 11-point sequential acceleration. Interestingly, while Q2’s guidance was nearly $4 billion ahead of the consensus estimate for 0.9% growth to $25.26 billion in revenue, Dell opted to maintain its FY26 revenue forecast at $101 to 105 billion.

AI Segment Revenue:

Dell reported surging demand in AI optimized servers in Q1 with orders of $12.1 billion. This outpaces the entirety of last year while representing a 612% sequential increase from $1.7B last quarter. To further compare, the peak quarter for orders last year was $3.6B.

This surge in orders brought Dell’s AI server backlog up to $14.4 billion, up from $4.1 billion in Q4. However, Q1’s AI server shipments were just $1.8 billion, up just 6% YoY and down more than (14%) QoQ. This likely boils down to the timing of Blackwell’s ramp, as Dell projected more than $7 billion in shipments in Q2

This strong AI server shipment forecast contributed to a nearly $4 billion beat for Q2’s guidance. Notably, Dell did not raise its revenue forecast for the year, suggesting that tariff-related impacts may still bite in H2, or that AI server shipments will be lumpy and not be linear from here out.

Earnings:

For Q2, Dell guided for $2.15 to $2.35 in adjusted EPS for growth of 15% at midpoint, marking a slight deceleration from the 17.4% growth reported in Q1.

Q3 and Q4 are expected to see EPS growth decelerate a bit further, with growth of just 10.7% in Q4.

For the full year, Dell slightly raised its FY26 adjusted EPS guidance to $9.40 for 15% growth, up from its prior view for $9.30 for 14% growth.

Dell also slightly hiked its GAAP EPS view for FY26, now seeing $7.99 for 25% growth versus its prior view of $7.85 for 23% growth.

Margins:

GM = 21.1%
GAAP Operating Margin = 5%
Net Margin = 4.1%

Margins are decel'ing which is an issue since AI servers weigh on margins. However, management expects to see ISG improve by $0.5 billion in operating income this quarter on an additional $5.3 billion in revenue – meaning management is sensitive to the importance of operating efficiency.

Cash flow:

Dell is reporting strong cash flow growth – setting itself apart from Supermicro:

Operating cash flow rose 168% YoY to $2.80 billion. OCF margin was 12.0%, up more than 7 points from 4.7% a year ago and more than 9.5 points higher than Q4’ s 2.4% margin.

Free cash flow rose 388% YoY to $2.23 billion, while adjusted free cash flow rose 258% YoY to $2.23 billion. FCF and adjusted FCF margin was 9.5%, a significant improvement from 2.1% and 2.8% a year ago.

Cash, equivalents and investments totaled $9.29 billion, up more than $4 billion QoQ. Debt also rose more than $4 billion QoQ to $28.78 billion.

Valuation:

Dell trades at 0.82 fwd PS and 13.4 fwd PE Ratio. This is at the medium-range of the company’s recent stock history since the AI boom began.

Notable Risks:

Dell is exposed to lower-performing Client markets, which equal higher revenue than its AI segment (ISG). Notably, ISG will likely overtake Client sometime this year in total revenue. Dell’s margins are very low

5. Amphenol: Leading AI Supplier with 134% Growth

Amphenol plays an important role in Nvidia’s NVL72 racks that are shipping now, as the company supplies high-speed copper cables and interconnects. Nvidia’s choice to use copper cabling over optical transceivers resulted in both lower costs and power savings for the NVL72, providing a growth opportunity for Amphenol. Specifically, Amphenol's 12VHPWR PCIe 5.0 power connector was able to eliminate the need for three power connectors with a single power connector.

Unlike other GPU-agnostic players who can realize growth and tailwinds as long as AI capex remains strong, Amphenol is more closely correlated to Nvidia’s NVL72, and its opportunity thus arises squarely from the ramp of the platform and overall shipment volumes. Signs that Nvidia is now quickly ramping NVL72 shipments far ahead of analyst expectations support more growth ahead for Amphenol in the upcoming quarters.

Amphenol’s dollar content per NVL72 rack is expected to be quite high — Evercore ISI estimated last year that Amphenol’s BOM content was in the range of $100,000 to $120,000 per NVL72, or around 3-4% of the server’s value. This represents a fairly large opportunity for Amphenol, especially if Nvidia is scaling shipments to a much larger degree than currently anticipated.

However, Amphenol remains quite highly exposed to slower-moving sectors such as the industrial and automotive sectors, and cash to debt is upside-down due to its focus on M&A to complement growth.

Overall Revenue:

Accelerating AI demand drove Q1’s outperformance, with revenue coming in “much stronger than expected” at 47.7% YoY to $4.81 billion in revenue, accelerating 18 points sequentially.

Organic revenue growth was 33%, accelerating 13 points sequentially. Q2’s growth is now expected to be 38.8%, more than 21 points higher than January’s 17.5% estimate.

Q3’s growth is expected to be 27.0%, approximately 16.5 points higher than in January.

Q4’s growth is expected to be 21.5%, nearly 10 points higher than in January.

AI Related Revenue:

134% in Datacom IT. Amphenol’s orders have grown at 58% YoY for a second consecutive quarter, with growth accelerating sharply over the last few quarters. Putting this together, the nature of Q1’s beat and the strength in datacom at 134% YoY has driven estimates for the next three quarters up by more than $2 billion combined.

Earnings:

Amphenol reported a quite large 21.2% beat on adjusted EPS in Q1, posting $0.63 versus the $0.52 estimate. This represented growth of 57.5% YoY, accelerating from 34.1% growth last quarter.

However, similar to revenue, growth is currently expected to peak in Q1 and decelerate after, though remaining quite strong. For 2025, Amphenol is expected to report 40.8% growth to $2.66 in adjusted EPS, with growth forecast to slow dramatically to the 9% range for both 2026 and 2027, in an indication that 2025 is expected to be the sole strong growth year for the company due to Blackwell’s initial ramp phase

Margins:

Gross Margin = 34%
Operating margin = 21.3%
Net Margin of 15.3%

Amphenol’s margins have been relatively stable over the past four quarters, but the strong growth and increasing contribution from Communications, which is accretive to operating margin, provides some margin tailwinds.

Cash:

Operating cash flow was $764.9 million for a 15.9% margin, down from 18.4% margin in the year ago quarter. OCF margin over the past three years has hovered between the 17% to 20% range, with Q1’s cash flow slightly weaker.

Free cash flow was $580.4 million for a 12.1% margin, down from a 15.5% margin in the year ago quarter. Management expects to have elevated capex again in Q2 to support datacom growth, weighing on FCF.

Valuation:

Amphenol’s valuation is stretched at 6 forward PS and 37 forward PE Ratio, some of the highest in the company’s history.

Risks:

Valuation is the primary risk as the company has a strong AI story yet overall revenue is low given the other, low-growth segments.

Read more in our analysis Amphenol Reports 134% Growth in Datacom IT SegmentAmphenol Reports 134% Growth in Datacom IT Segment

6. Coherent: Lesser-Known Supplier Reporting Inflection in AI-Related Revenue

Coherent is reporting Q4 fiscal year 2025 this quarter

Coherent reported a double beat in Q3 with revenue growth of 24% and EPS growth of 141% YoY. The top line beat was driven by Data Center and Communications revenue growing 46% YoY. While this growth moderated slightly from the prior quarter, Nvidia suppliers should see a meaningful acceleration in the second half of the year.

Analysts have yet to fully factor in this acceleration, but as NVIDIA ramps Blackwell-based systems and scales out its Spectrum-X Ethernet and Quantum-X Infiniband platforms, suppliers of high-speed optical interconnects are likely to see an increase in demand. Coherent, as a key ecosystem partner to NVIDIA in silicon photonics and co-packaged optics (CPO), is well positioned to benefit as hyperscalers upgrade to 800G, 1.6T, and eventually 3.2T.

To refresh your memory, Coherent has many products that participate in the AI-driven datacom transceiver and optical interconnects market. Primarily, the growth story centers around supplying Nvidia with pluggable optical transceivers (400G, 800G, 1.6T) including EML lasers, VSCEL lasers and CW lasers, and emerging CPO technologies for next-generation switches and interconnects.

Coherent is certainly not without competitors, and this is the main risk the company faces. Management is tasked with executing flawlessly in an environment where components may see supply disruptions and must also move quickly to make sure they are first to market to support higher bandwidths. Optical transceivers are at risk of being commoditized as reflected in Coherent’s low margins.

Overall Revenue:

Coherent reported a double beat in Q3 with revenue growth of 24% and EPS growth of 141% YoY. Coherent delivered another quarter of record revenue driven by strong AI data center demand, with revenue rising 4.4% QoQ and 23.9% YoY to $1.50 billion.

This beat the consensus estimate for $1.44 billion by more than 4%, and marks a third straight quarter of >20% revenue growth. For Q4, management guided a wide range for revenue, forecasting $1.425 to $1.575 billion.

At the $1.5 billion midpoint, this represents flat QoQ and 14.5% YoY growth, slightly ahead of estimates for 12.1% growth. Revenue growth estimates for the next two quarters have moved higher since our last Q2 report, from the mid-9% range to double-digit growth through FQ1 2026.

AI Segment Revenue Growth:

The top line beat was driven by Data Center and Communications revenue growing 46% YoY. Networking revenue increased 46% YoY and 10% QoQ to $897 million, or ~60% of revenue.

Notably, growth continues to decelerate from Q1’s 61% print, yet the segment’s growth is much stronger this year compared to last. For the first nine months, networking revenue was $2.48 billion, up 53% YoY.

According to a press release in March, Coherent was the first to release a 400G per lane EML for 1.6T, showing Coherent is working hard to remain a supplier of choice in a highly competitive market. To some extent, indium phosphide capacity is the limiting factor for these technologies, with Coherent stating they expanded capacity rapidly in the current quarter: “In Q3, we once again expanded our capacity both sequentially and year-over-year with year-over-year capacity growing by over 3x.”

Earnings:

Coherent reported a 5.8% EPS beat in Q3 as it benefited from strong margins down the line, reporting $0.91 in EPS. This represented growth of 141% YoY, decelerating from 256% YoY growth in Q1.

For Q4, management offered a wide range for $0.81 to $1.01 in adjusted EPS, with the $0.91 midpoint in-line with estimates. For FY25, Coherent is currently expected to record more than 107% YoY growth to $3.46, though growth is expected to slow to 26.2% YoY to $4.37 in FY26.

Margins:

Q3 GAAP gross margin was 35.2%, expanding nearly 5 points YoY.
GAAP operating margin was 4.8%, up 3 points YoY.

For Q4, management is holding adjusted gross margin guidance steady at 37-39%, while guiding for an 18% adjusted operating margin. Coherent is beginning to close in on its long-term gross margin targets of 40% over the last two quarters, though it still needs to make some considerable progress or drive faster growth in higher-margin products to reach this threshold in fiscal 2026.

Note: Coherent is expected to divest low-margin segments soon which would quickly change its margin profile.

Per our previous writeup:

“The company recently restructured the business to divest the silicon carbide portion, which is also contributing to better margins for next quarter: “So I think you're referring to some of the restructuring that we've taken and the portfolio actions associated with it. And so what I would say is that the actions that were taken in terms of an underutilized assets or underutilized businesses, that benefit is — certainly will contribute to our financials from a gross margin and OpEx perspective, depending on the nature of the actual divestiture.”

Cash:

Operating cash flow was $162.9 million for a 10.9% margin, expanding from a 9.7% margin a year ago. This was the fourth consecutive quarter of a double-digit OCF margin.

Free cash flow was $51.1 million for a 3.4% margin, expanding from a 2% margin a year ago

Valuation:

Similar to Lumentum, Coherent shows room in its bottom line valuation whereas there is less room in the top line valuation. At 21 forward PE ratio, the company is trading at its lowest in two years. At a 2.3 forward PS ratio, it’s closer to the top valuation its traded at in two years at 3 – which seems to be a firm ceiling unless there is a re-rating on the AI story.

Notable Risks:

There are a few competitors Coherent must contend with, its lower-growth segments weigh on the stock. The margins leave a lot to be desired.

Honorable Mentions:

Lesser-known supplier at inflection point, covered on Discovery tier April 29th

We covered a lesser-known supplier that offers components for datacom transceivers and optical interconnects on April 29th. This small-cap company offers differentiated technology that has caught the attention of heavyweight NVIDIA. We’ve been patiently waiting for this company’s EML lasers for 200G to ship, enabling 800G and 1.6T bandwidths. Any progress here will continue into 2026-2027 for 400G data lanes and 3.2T bandwidths.

Read more about our Discovery tier here.

Thermal Management Solutions Provider, covered on Discovery on May 15th

We recently covered a supplier whose importance is expected to increase with each new generation of GPUs and AI accelerators. The company provides thermal management solutions, such as cold plate cooling and immersion cooling to lower the power requirements to AI systems. They also offer high density solutions such as rear door heat exchangers and coolant distribution units (CDUs).

Direct liquid cooling systems, including hybrid versions that combine air and DLC, can result in 40% less power management space and 20% lower cooling costs. When you’re spending nearly $100 billion per year on capex like many Big Tech companies, this matters quite a bit. In addition to thermal management, this company's power solutions include uninterruptable power systems and lithium-ion battery cabinets that supply up to 1500KW and 263KW in a single cabinet.

Read more about our Discovery tier here.

Section 3: AI Software – Strong Fundamentals Yet Valuations are Stretched

There are six stocks in this section ranked #1 to #6 and two honorable mentions

AI software valuations are pointing toward a bubble within a larger, quality trend. We are simply too early for AI software to carry the valuations some are commanding and the evidence is quite clear in the financials when you compare valuations to AI hardware.

AI hardware segments are often growing triple-digits+ and are likely to continue to do so for some time
AI software segments have either dipped below 100% (with major hurdles to resume this growth) or have not surpassed 100% — and yet the valuations are in some cases Covid-era like.

The stocks below make the list but will only make the final cut (added to the portfolio) if we can get the stocks at a reasonable valuation – which is true for the entire list, but there is nothing quite like getting a solid software stock at a reasonable valuation. That is the trade that rules all trades in the tech sector.

1. AppLovin has a Rule of 140 (not a typo)

Thematic: 10/10
Fundamentals: 10/10
Valuation: 2/10

AppLovin easily topped revenue and EPS estimates in Q1, but more importantly, the company is setting up for an additional under-reported catalyst with its web-based ad platform expected to launch its self-serve feature and scale with a wider pool of advertisers as the year progresses.

In addition, the company divested its App segment, which is the gaming assets portfolio, and is now a pureplay ad-tech stock. The high-growth and high-margin advertising business that ignited AppLovin’s strong returns over the past few years is now the company’s sole focus.

You’d be hard pressed to find a stronger stock in terms of fundamentals on the market today. There is plenty of runway left for this stock should the growth of 30%+ coupled with 80%+ gross margins and nearly 40% net margin continue. Consider that EPS grew triple digits this quarter and FY2026 EPS estimates are being revised higher by an astonishing $3.50 in incremental EPS.

Overall Revenue Growth:

AppLovin reported 40.3% YoY revenue growth to $1.48 billion in the first quarter, beating consensus estimates by $100 million. This was AppLovin’s sixth consecutive quarter with revenue growth >35% YoY.

Given the Apps business is being divested, AppLovin will be reporting headline growth in the 60% range that is aligned with its Ads business rather than a mix of both. Consensus revenue growth estimates are much lower and show a sharp deceleration, as these comps still take into account revenue from the Apps segment. Thus, growth rates such as 20% in Q3 do not reflect the true performance of the business.

AI Segment Growth:

Advertising revenue increased 70.9% YoY to $1.16 billion, slowing slightly from 91% in the year-ago quarter. Management said growth was driven by continued enhancements in its AI ad engine, as well as the full quarter impact of its web-based ad solution even coming off the seasonally high e-commerce quarter in Q4.

For Q2, management guided Advertising revenue of $1.195 to $1.215 billion, pointing to 69.5% YoY growth at midpoint, maintaining its hypergrowth phase.

Earnings:

AppLovin’s business model sees a high percentage of its operating income flow through to the bottom line, driving tremendous EPS growth as margins expand.

AppLovin reported massive 149% YoY growth in GAAP EPS to $1.67, outpacing revenue growth by more than 3x.

Q2 EPS is now seen growing 125% YoY to $2.00, before rising to $2.16 in Q3 and exiting the year at $2.46. For FY25, analysts estimate AppLovin will generate $7.80 in EPS, up 72.3% YoY, with FY26 EPS rising 42% to $11.80. This is more than a $3.50 increase for FY26 since February’s $8.27 estimate.

Margins:

Though AppLovin’s top-line growth is quite impressive, margins are where it shines, with gross margin surpassing 80% and operating margin reaching a new high. This combination of strong revenue growth and strong margins is driving exceptional operating leverage with triple-digit earnings growth.

Gross margin expanded 5 points sequentially and more than 9 points YoY to 81.7%. Notably, AppLovin cut its cost of revenue by nearly (9%) YoY, from $294.1 million to $272.2 million, while still driving 40% total revenue growth and 70% advertising growth.
Operating margin remained above 44% for a third straight quarter at 44.7%. To put in perspective how strong these margins are, AppLovin would have a Rule of 40 score of 85% based on Palantir’s definition of revenue growth + operating margin, while Palantir had a score of 83%.
Net margin in Q1 was 38.8%, up more than 16 points YoY.

However, now that the Apps business is divested, the operating margin will skyrocket to 70% range for a Rule of 140 if you assume 70% revenue growth on the Ads business and 70% operating margin. This is unheard of; I do not think we’ve seen this combination before of such high growth and such high profitability. Typically, software startups and public companies are seeking a Rule of 40 and yet AppLovin offers 100-points higher following divesture of the Apps segment.

Cash:

AppLovin’s cash flows are exceptional, with operating and free cash flow margins expanding to new records in Q1. Per the opening remarks: “In the first quarter, we generated $826 million in free cash flow, up a staggering 113% year-over-year. Quarter-over-quarter, our free cash flow grew 19%, representing an impressive 82% flow-through from adjusted EBITDA to free cash flow.”

Operating cash flow rose 112% YoY to $831.7 million for a 56% margin, expanding from a 51.1% margin in Q4 and nearly 19 points higher than 37.1% in the year ago quarter.
Free cash flow rose 113% YoY to $825.7 million for a 55.6% margin, expanding from 50.6% in Q4 and 36.6% in the year ago quarter.

Valuation:

Applovin is trading at 22 forward PS which is at the upper range for this stock, topping at 30 two times (briefly) in the past before retreating as low as 15. The PE Ratio of 40 is similar as it’s well above 3-year and 5-year medians while seeing a brief top at 60-70 before quickly retreating to a low of 27.

Risks:

Of the software stocks, Applovin has fewer risks than the other stocks given its valuation is in typical range. The divesting of the Apps business is bullish but results in tough comps for headline numbers. Investors will want to focus on organic growth the ads business.

You can read more about Applovin in the analysis: “AppLovin Q1: Web-Based Catalyst 2025-2026; App Segment Divested is a Major Plus”AppLovin Q1: Web-Based Catalyst 2025-2026; App Segment Divested is a Major Plus”

2. Oracle: The AI Software Stock No One Saw Coming

Oracle laid out some impressive growth forecasts for fiscal 2026 earlier in June, setting the stage for a significant acceleration in its cloud segment backed by robust AI demand. The recent 4.5 gigawatt agreement with OpenAI for expanded Stargate capacity is a testament to Oracle's aggressive push in the AI cloud market, strengthened by its focus on low-latency, high performance AI.

This massive deal, requiring significant data center expansion, underscores just how elevated demand is for high-performance infrastructure to power advanced AI models. Analysts are closely watching how this mega-deal impacts Oracle's capex strategy and its overall AI growth outlook for the coming fiscal years.

Notably, Cloud IaaS (OCI) growth was guided to accelerate to >70% in FY26 from 50% in FY25. Oracle said this acceleration was supported by “exceptional demand infrastructure services” and non-cancelable RPO bookings.

Overall Revenue Growth:

As a result of the strong forecasted growth in cloud and in RPO, Oracle slightly raised its FY26 revenue target, while management stated they have increased confidence in meeting and possibly exceeding FY27 and FY29 targets.

For FY26, Oracle now expects revenue to be $67 billion for YoY growth of 16.7%, up just $1 billion from its prior guidance for $66 billion. This slight raise corresponds to a 1.7 point topline acceleration, from barely 15% YoY in its original forecast to 16.7% now with the potential for cloud-driven upside now that GPUs are no longer a constraint.

However, the small increase raises some questions about the durability of Oracle’s non-cloud growth given the magnitude of acceleration management sees in the cloud.

Put it this way – if cloud was previously expected to accelerate nine points to 33% YoY ($32.4 billion) in FY26, the new 16 point-plus acceleration guide would raise cloud revenue $1.8 billion higher to $34.2 billion. Thus, the $1 billion full-year hike suggests that Oracle’s non-cloud segments may be flat at best, or decline low-single digits YoY.

AI Revenue Growth:

One of the more impressive forecasts Oracle stated in Q4 was its RPO growth target for FY26. Management stated that RPO was expected to increase more than 100% in the upcoming fiscal year, which would place RPO at well over one-quarter trillion.

During its Q4 report in early June, Oracle projected substantial acceleration in its cloud business in fiscal 2026, fueled by strong AI demand and cloud consumption:

Total cloud growth (IaaS & SaaS) guided to accelerate to >40% in FY26 from 24% in FY25.
Cloud IaaS (OCI) growth guided to accelerate to >70% in FY26 from 50% in FY25. Oracle said this acceleration was supported by “exceptional demand infrastructure services” and non-cancelable RPO bookings.
Oracle Cloud Infrastructure consumption revenue to grow faster than the 62% YoY increase reported in Q4.

In terms of revenue, these growth rates project total cloud revenue rising to at least $34.2 billion, up from $24.4 billion in FY25. Cloud IaaS is projected to rise to at least $17.5 billion increasing from $10.2 billion in FY25. In terms of revenue mix, IaaS would see its share rise quite sharply, from 41% of cloud revenue in FY25 to >50% in FY26.

Oracle’s IaaS growth segment is expected to increase 4.5x by FY28, with the $30 billion deal then kicking in.

As stated above, consensus currently models in $46 billion in IaaS revenue in FY28. For the IaaS segment to increase 4.5x from FY25’s $10.2 billion in revenue, this requires growth at a 65.2% CAGR, or a slight deceleration from >70% YoY in FY26 to >60% YoY in both FY27 and FY28.

Though Oracle is growing off a much smaller cloud base than, say Azure, this represents potentially a 30 point faster growth rate than its hyperscaler peers over the next few years. It’s also representative of a significant reshaping of Oracle’s business model, as this implies IaaS will grow its share of cloud revenue from <18% to nearly 50% in just three years.

The $30 billion annual revenue deal unlocks further upside in FY28 and into FY29, depending on how capacity and revenue ramp. It’s likely to take a couple years for Oracle to scale into the full run rate of the deal, but an additional $5B+ by FY29 could help Oracle easily exceed its targets on persistent cloud momentum.

Earnings:

Oracle reported a nearly 4% beat on the bottom-line, reporting $1.70 in adjusted EPS, rising just 5% YoY. For fiscal 2025, adjusted EPS rose 8% YoY to $6.03.

Adjusted EPS growth is projected to accelerate in fiscal 2026, driven by the top-line acceleration. Q1 EPS is estimated to increase 6.3% YoY and gradually accelerate to 17% YoY by Q4. However, full-year EPS growth estimates of 12.3% lag revenue growth by nearly 4 points, suggesting some margin headwinds may be present throughout the year.

Margins:

Oracle’s margins are solid and have remained quite steady, with only marginal expansion down the line.

FY25 GAAP gross margin was 71%, flat YoY, though gross margin in Q4 was down 3 points YoY to 70%.
FY25 GAAP operating margin expanded 2 points YoY to 41%, strengthening slightly throughout the year from 30% in Q1 to 32% by Q4.
FY25 adjusted operating margin was flat YoY at 44%.
FY25 GAAP net margin followed operating margin, expanding 2 points YoY to 22%.
FY25 adjusted net margin was flat YoY at 30%.

Cash:

Oracle’s liquidity profile suggests that funding aggressive expansion plans, notably for Stargate, will pressure free cash flow through 2026 and potentially into 2027 as well. Capex outpaced operating cash flow for the first time in FY25, sending FCF negative.

Cash, equivalents and short-term investments totaled $10.2 billion, though debt was 9x this at $92.6 billion.
Operating cash flow in Q4 was nearly $6.2 billion for a 39% margin, though free cash flow was more than ($2.9 billion), or an (18%) margin. This is because capex rose 55% QoQ to $9.1 billion.
For FY25, operating cash flow was $20.8 billion for a 36% margin, up 1 point YoY. Free cash flow was ($0.4 billion), for a (1%) margin, down from a 22% margin in FY24 due to Oracle spending $21.2 billion in capex.
For FY26, Oracle guided for capex of >$25 billion, and hinted that actual requirements could be higher, suggesting FCF may be negative again unless operating cash flow growth accelerates from 12% to >25%.

Valuation:

Oracle is trading at 10 forward PS, which is not a valuation we’ve seen from this stock. Typically, it trades in the 5X forward range. It’s anyone’s guess if the company will join AI peers at higher valuations or retreat back to its typical valuation.

The PE Ratio is the same – its 2X higher compared to historic levels at 35 forward PE and we will need more AI-related trading history to determine where the stock eventually settles.

Risks:

No major risks. Valuation could go either way therefore less of a risk and more of a trial period.

You can read more about Oracle in the analysis: “Can Oracle Become the Next $1 Trillion AI Stock?”Can Oracle Become the Next $1 Trillion AI Stock?”

3. Palantir is the AI Bubble Stock

Thematic: 7/10
Fundamentals: 10/10
Valuation: 0/10

Palantir is at an eye-watering valuation – causing many investors with no risk management to come out of the woodwork and cheer the stock at these levels. That may work in many sectors, but it does not work in the tech sector. It would take a significant selloff for our firm to buy Palantir right now as we simply refuse to gamble with our hard-earned money. Don’t hate me, but I actually like Oracle better here than I do Palantir.

If you give me Palantir at a drastically better valuation that matches what best-of-breed cloud stocks can sustain, then I’d happily buy. Until then, we have our attention on other stocks for now.

Overall Revenue Growth:

In Q1, the company reported $884 million in revenue for growth of 39%, up from growth of 36% last quarter and 21% last year. This represents QoQ growth of 7%.

Palantir reported $883.9 million in revenue in Q1, beating estimates by more than $21 million. As stated above, this represents growth of 39%, up from growth of 36% last quarter and up from 21% last year.

On a QoQ basis, Q1 accelerated 7% from Q4. This is an impressive performance given Q1 is typically one of the slowest quarters seasonally.

AI Related Revenue Growth:

Perhaps most importantly, US commercial revenue drove the results, with 71% YoY growth and QoQ growth of 19% for the segment’s first-ever $1 billion annual run rate.

US commercial revenue accelerated from 64% last quarter to 71% YoY this quarter to $255 million, surpassing a $1 billion annualized run rate for the first time on elevated AI demand.

However, the guide for next quarter does indicate Q1 could be the peak with fiscal year growth of 68% guided. Palantir raised its FY25 US commercial growth guidance from 54% YoY to 68% YoY, projecting revenue of $1.178 billion, compared to $457 million in 2023. The raise represents about $100M more than previously expected.

Earnings:

Despite the top-line beat, Palantir met adjusted EPS estimates in the quarter at $0.13, up 68% YoY. GAAP EPS was $0.08, up 100% YoY.

Looking ahead through the rest of FY25, adjusted EPS growth is expected to decelerate, from Q1’s 68% YoY to 20% YoY by Q4. However, estimates have risen over the past three months – Q2’s growth rate has come up 11 points and Q3’s up by 9 points.

For FY25, Palantir is expected to see adjusted EPS growth of nearly 43% YoY to $0.58, before decelerating to 25% growth to $0.73 in FY26.

Margins:

Across the board, Palantir has been expanding its margins. Adjusted EBITDA margin was 45% — which is one of the highest in the tech universe.

GAAP gross margin was 80.4% in Q1, down 1.3 points YoY.
Adjusted gross margin was 82.1%, down more than 1 point YoY.
GAAP operating margin expanded to 19.9%, up more than 7 points YoY.
Adjusted operating margin was 44.2%, up 8.5 points YoY. For Q2, Palantir guided its adjusted operating margin to 43.1%, which would represent a third consecutive quarter above 40% and up nearly 6 points YoY.
GAAP net margin was 24.2%, up more than 7.5 points YoY.
Adjusted net margin was 37.8%, up nearly 8 points YoY.

Cash:

Palantir stands out for its ridiculously strong cash flows, though operating and free cash flow margins moderated quite substantially in Q1 relative to 2H 2024.

Operating cash flow was $310.3 million in Q1 for a margin of 35%, down from 56% in Q4.
Adjusted free cash flow was $370.4 million for a 42% margin, down from a 63% margin in Q4. Palantir raised its adjusted FCF guidance for FY25 from $1.5-1.7 billion to $1.6-1.8 billion, implying an FCF margin of 43.7%.

Cash and equivalents totaled $5.43 billion, while debt was zero.

Valuation:

90 forward PS is a valuation we have not seen since Snowflake traded after its IPO. Today, Snowflake trades at 15 forward PS. Yes, AI deserves a premium. However, at 90X sales means Palantir will not pay back the valuation in revenue in this lifetime as you’re paying a valuation worth 90 years of revenue – and that’s assuming Palantir maintains its current growth rate.

Notable risks:

Valuation, valuation, valuation

You can read more about Palantir in the analysis: “Palantir Stock: Strong Sequential Growth and Strong Underlying Key Metrics.”Palantir Stock: Strong Sequential Growth and Strong Underlying Key Metrics.”

4. Cloudflare: The Upcoming AI Inference Darling

Act 3 refers to the Workers platform, which is the company’s attempt to compete with hyperscalers – but most importantly, it sets up the company well for AI inference at the edge.

When it comes to AI inference-driven revenue, it’s still relatively early in the growth curve. Hyperscalers and model providers only recently began to disclose rapid AI token growth over the last three to four months. However, Q1’s earnings report shows signs of surging AI inference demand filtering into Cloudflare’s platform. For example, Q1 witnessed nearly 4,000% YoY growth in Workers AI inference requests, and more than 1,200% YoY growth in AI Gateway requests.

These growth numbers are off a small base (which is true for all inference statistics for now), yet when you take a company with product-market like Cloudflare and combine it with a massive trend on the verge of taking off – what you get is an irresistible stock that the I/O Fund has a high probability of entering and holding for an extended period of time.

Overall Revenue Growth:

Cloudflare reported a 2% beat in Q1 with revenue increasing 26.5% YoY to $479.1 million. This growth was attributed to the strength of Cloudflare’s largest >$1M and >$5M ARR customer cohorts, which saw record customer additions in the quarter.

Looking ahead to Q2, Cloudflare guided for 24.8% YoY growth to $500 million to $501 million in revenue, representing a 1.7 point sequential deceleration. Analysts are much more optimistic on the quarter, projecting growth above the top end of the range at $501.8 million, or up 25.1% YoY.

Through the rest of fiscal 2025, growth is expected to be essentially flat around 25% YoY. However, management expressed confidence in driving a reacceleration through 2025, opening the door for potential upward surprises driven by AI inference.

However, by maintaining guidance despite the $10M beat in Q1, Cloudflare is essentially saying Q2 could be softer than expected. With that said, Cloudflare tends to be conservative during macro events such as what we saw in April, and thus it could also be a non-issue.

AI Segment Growth:

Cloudflare does not break out its AI segment too closely, rather they share initial growth numbers on the Workers platform. Per our intro: “Q1’s earnings report shows signs of surging AI inference demand filtering into Cloudflare’s platform. For example, Q1 witnessed nearly 4,000% YoY growth in Workers AI inference requests, and more than 1,200% YoY growth in AI Gateway requests.”

Additionally, there are some key metrics that seem to have bottomed and are finding a tiny inflection point:

In Q1, paid customer growth accelerated 2 points sequentially to over 27% YoY, with Cloudflare reporting 250,819 paid customers. Growth has doubled from 13% two years ago, an impressive acceleration given the scale is now reaching a quarter-million paid customers.

Cloudflare also noted it had driven record customer additions in its >$1M and >$5M ARR cohorts in Q1, with growth in both metrics up 48% and 54% YoY, respectively.

Billings growth also accelerated 1 point to 32.8% YoY in Q1, recovering from the 20% range in 2024. Billings activity likely benefitted from QoQ improvements in sales cycles as noted in Q1, as well as stronger deal activity and larger contracts.

Cloudflare’s DBNRR stabilized at 111%, though it has yet to see a strong acceleration like Palantir. Compared to last year, DBNRR is 4 points lower.

RPO also reaccelerated in Q1 to nearly 39% YoY to $1.86 billion, though there has been consistent quarterly variability in growth over the last two years. Current RPO accounted for 66% of total RPO, down from 70% in Q4.

Earnings:

Cloudflare reported adjusted EPS in line with estimates at $0.16 in Q1, for flat YoY growth. Q2 is expected to see adjusted EPS decline mid-single digits YoY, with the full-year on track for just mid-single digit growth with an acceleration expected in Q4.

For Q2, Cloudflare guided for adjusted EPS of $0.18, down from $0.20 in the year ago quarter. Adjusted EPS growth is expected to resume in 2H, with EPS seen exiting the year at $0.23, up 22.6% YoY.

For FY25, Cloudflare maintained its guidance for $0.79 to $0.80, corresponding to growth of approximately 6% YoY. For FY26, analysts are projecting EPS growth to accelerate sharply to 30.3% YoY to $1.04, which likely would require solid improvement in adjusted margins given the topline acceleration is minimal.

Margins:

Margins are the one real blemish for Cloudflare, as the company has regressed on its path to reach GAAP profitability in Q1. Gross margins have been compressing slightly, due to an increase in paid versus free traffic, while operating margins slipped sequentially in Q1.

GAAP gross margin was 75.9% in Q1, down 0.5 points sequentially and 1.6 points YoY. Adjusted gross margin was 77.1%, down 0.5 points sequentially and 2.4 points YoY.
GAAP operating margin was (11.1%) in Q1, down 3.6 points sequentially and a setback from three consecutive quarters of progress towards profitability in the (7%) to (8%) range.
Adjusted operating margin was 11.7%, marginally above guidance for 11.6% and down 2.9 points sequentially. For Q2, Cloudflare guided for adjusted operating margin to improve one point to 12.6%.
GAAP net margin was (8.0%) in Q1, a rather substantial decline from (2.8%) last quarter, driven by the QoQ decline in operating margin. Adjusted net margin was 12.2%, the lowest reported level since Q2 2023.

There is nearly a 23 point gap between GAAP and operating margins. This is driven primarily by high SBC at ~20% of revenue, and it highlights that either SBC would need to move much lower, or costs much lower, in order to drive Cloudflare to a sustainable path to GAAP profitability.

Cash:

Operating cash flow continues to improve, touching a 30% margin in Q1, though free cash flows remain pressured by heightened network capex at 17% of revenue. Cloudflare also raised a substantial amount of capital on June 13, an interesting move given the company still has nearly $2 billion in cash on hand.

Operating cash flow rose more than 98% YoY to $145.8 million, for a 30% margin. This marked a substantial 11 point improvement from a 19% margin a year ago and a 2 point sequential improvement.
However, free cash flow rose 48.6% YoY to $52.6 million, for an 11% margin, up only 2 points YoY.
Cash and investments totaled $1.92 billion, while Cloudflare reported $1.29 billion in convertible debt still outstanding, due in 2026

Valuation:

Cloudflare’s Forward PE ratio is wild at 234 compared and its forward PS ratio of 31 is also a bit steep.

Notable Risks:

GAAP operating margin is weak. Keep an eye on capex costs.

This stock also carries execution risk in both directions as trying to time Cloudflare’s big moment will take immense skill. An investor could buy now and wait … or put that money to work elsewhere and return when there’s more indication that startups, SMBs and enterprises are willing to pay for edge inference. With what we know today, the AI market is primarily driven by Big Tech (i.e., not customers of Cloudflare). Palantir is a great example, it’s finally at a $1B annual run rate and is considered by many the leading AI software stock. Therefore, paying 31 forward PS and risking Cloudflare returns to 15 forward PS (it’s typical range) is not only a valuation risk, but also an execution risk in terms of how long you’d have to hold the stock to return to previous levels, as unlike Palantir, Cloudflare is not showing material AI revenue (yet).

For more information on connecting the dots for Cloudflare’s AI inference thesis, reference our analysis: “Cloudflare: Entering Act 3 to Become A Leader in AI Inference at the Edge.”Cloudflare: Entering Act 3 to Become A Leader in AI Inference at the Edge.”

5. Small Cap Stock with up to 135% Growth, Undervalued Relative to Opportunity

Our team has recently covered a small cap stock on the Discovery tier that we believe could ultimately become a sizable winner. Find more details on our Discovery tier here.

Our team works hard to dig up new ideas, which we publish on the Discovery tier, and this is one of the team’s favorites over the past 2-3 months. Knox has two setups outlined, and they both indicate this stock will see a drop before (potentially) becoming one of our highest performers. Our plan is to enter on any weakness or on a meaningful breakout.

General Synopsis:

Complex reasoning models require an expanded data set, such as dozens of foreign languages or multi-step problems within math and chemistry, for example. This is in contrast to a static data set, which often produces too many hallucinations and can be inaccurate at times. For example, if a Big Tech company only used its proprietary social data to train LLMs, this may not be broad enough to prevent hallucinations since social data is limited in its context and scope. In many cases, additional data points are sought out to improve the accuracy of the model.

In order to move toward general artificial intelligence (AGI), which is defined as AI models that think for themselves similar to a human, companies like Innodata are also tapped for their ability to augment accuracy through reinforcement learning and direct preference optimization, which utilizes subject matter experts to annotate data and to also stress-test the models for accuracy.

This company's competitor is valued at $29 billion compared to the I/O Fund's stock pick having a market cap at $1.5 billion on last year’s reported revenue of $870 million last year. If we assume the competitor is at $1 billion revenue now, that would be a 29X compared to our pick's 6X forward sales.

Speaking of said competitor, there is a potential exodus from the competitor as they received a large funding round from a Big Tech company and this is seen as potential IP risk by other Big Tech companies that previously used the competitor for labeling data sets. This could become a windfall for the I/O Fund’s stock pick. Sign up for Discovery to get our stock trades on this small cap in addition to ongoing coverage.

Current Pro and Advanced Members: To subscribe to Discovery with 30% off, please click here to email usclick here to email us or email premium@io-fund.com and mention code DISCOVERY30.

Discovery is aimed at surfacing new ideas. Rather than being confined to the I/O Fund’s portfolio coverage, Discovery unleashes new ideas to be early with the goal of providing significant edge to tech investors.

Honorable Mentions:

I’m calling these two honorable mentions not for lack of a strong thesis but because it’s a bit lame to include very well-known Mag 7 and FAANG stocks in a Top 15 list. You know these names well, and from here, we will do what we can to help you get a good entry. As of now, valuations are pretty stretched with Meta at the highest levels in the stock’s history.

Microsoft: The Undeniable AI Enterprise Juggernaut
If Nvidia holds the crown in the AI hardware arena, then Microsoft holds the crown in the AI enterprise arena.

Last quarter, Microsoft Azure was the only cloud provider of the three platforms to see growth accelerate, highlighting Microsoft’s impressive earnings for Q3 2025. Not only did Azure separate itself with this 4-point sequential growth acceleration, but it also grew at more than 2x the rate of AWS and 7 points faster than Google Cloud, reaffirming the company’s momentum in the Azure vs AWS vs Google Cloud battle.

Azure benefited as Microsoft brought capacity online faster than expected last quarter, to meet high demand for AI services. AI contributed 16 points of growth in the quarter, compared to 13 points last quarter and 10 points of growth a year ago. Microsoft did not provide an update on AI’s run rate yet said last quarter it had surpassed $13 billion, up 175% YoY.

Valuation:

After Microsoft’s fiscal year adjustment on July 1st, the stock is now trading at 33 forward PE implying at most a 10% move and its forward PS ratio is at 12. The stock typically tops at 13.5 max on Fwd PS and Fwd PE of 37 is a brief top before the stock retreats backward to as low as 23 to 27.

To read more about Microsoft’s recent quarter, including a few key points that are overlooked in terms of how the AI Enterprise juggernaut can extend its lead, reference our free article “What Separates Azure from AWS from Google Cloud” that is then continued on the premium side here.

Meta: Bottom Line Shines; Top Line Taking a Breather
Meta is supposedly no longer in the year of efficiency and is now in the year of AI, according to management. However, the efficiency was remarkable yet again last quarter. Although the company is decelerating from high growth in the past, the company has a big year ahead with ad improvements resulting in higher ad pricing, Meta AI standalone app recently launched (to be monetized next year), and its Llama 4 models, which are open source yet driving important productivity gains internally. Undoubtedly, the company has a lot of data for personalization and a highly engaged audience, marking two competitive advantages over other AI chatbots.

With that said, advertising key metrics decelerated sequentially, supporting further revenue growth deceleration for Q2. Ad impressions increased just 5% YoY, slowing considerably from 2023’s peaks and facing a tougher comp at 20% YoY last Q1. Ad pricing increased 10% YoY, a 4 point acceleration from 6% a year ago.

Valuation:

Meta is flashing warning signals with a Fwd PE ratio of 28 and a Fwd PS Ratio of 9.5. Two years ago, the stock traded at a forward PS ratio of 3 and a forward PE Ratio of 10. On a trailing twelve months basis, the stock is in line with historic trends, yet when a forward is decoupled from historic trends, it means investors are paying dearly for a stock with slowing growth.

Conclusion:

If you made it this far, congratulations! You must take your portfolio as seriously as we do. We have been heads-down attempting to squeak out higher returns this year and every inch of progress can make a big difference when it comes to positioning correctly.

Next up, Advanced Members will get technical setups from Knox in his Quarterly Positions Report with a complete picture of how we plan to enter the stocks listed above. Discovery Members will exclusively get updated technical setups on the three Discovery stocks that made the list.

Our results speak for themselves in terms of how we stack up, yet we continue to strive to move the needle on presenting to you the world’s best AI portfolio. Given the sheer ease in which the market moved off the April lows, we do foresee some volatility in the upcoming quarter. It’s nothing our team won’t be able to handle. Tech earnings officially kick off tomorrow – to say we are ready is an understatement. Let’s go!

Recommended Reading:

NuScale Power: Recent NRC Approval Paves Path to Commercialization

Posted on July 18, 2025June 30, 2026 by io-fund

NuScale is the only NRC-approved SMR module developer, holding a significant first-mover advantage in the space. The startup has one potential project underway in Romania with its majority stakeholder Fluor and RoPower, and a 1.85 GW agreement with Standard Power to power two future data centers in Pennsylvania and Ohio.

NuScale remains confident in its growth prospects, doubling modules in production to 12 as it aims to secure a firm customer order in 2025. Unlike Oklo, NuScale is partnering with energy investment and development firm ENTRA1 to sell and deploy its modules, and rough estimates imply each GW scale plant could represent several billions in revenue.

Despite this enormous opportunity, risks remain.

Gigawatt-Scale Projects Promise Multi-Billion Dollar Opportunities

As we discussed in our free newsletter last week, Nuclear Power Emerging as a Clean AI Data Center Energy Source, AI data center electricity demand is forecast to surge over the next few years. Most estimates from industry groups and analysts forecast 70 GW to 80 GW in demand growth through 2030, with focus turning to nuclear to meet long-term baseload energy needs.

Though some researchers estimate that at least 85 to 90 GW of nuclear energy is needed to help meet rising data center demand, high costs and lengthy construction timelines inhibit new large-scale deployments. Advanced microreactors and small-modular reactors (SMRs) promise a quicker path to deployment, though they’re also likely to take a small share of this demand growth. This is because companies in the space such as Oklo and NuScale have yet to commercialize modules and subsequently ramp production.

Regardless, even capturing just a small share of this demand growth, such as 5 to 10 GW over the next decade, SMR startups could see significant multi-billion dollar revenue opportunities.

NuScale recently received NRC approval at the end of May to increase the output of its module from 50 MWe to 77 MWe, offering flexible deployment options of four (VOYGR-4), six (VOYGR-6) or 12 module (VOYGR-12) plant designs. NuScale’s SMRs produce up to 462 MWe in the six-module plant up to 924 MWe in the 12-module plant.

NuScale has not been upfront about module pricing, though its original 2018 estimate with UAMPS for a 600 MWe, 12-module plant was ~$4.2 billion, or ~$350 million per 50 MWe module, before rising to $6.2 billion for an uprated 720 MWe plant in 2023, or ~$515 million per module.

This suggests that a ~1 GW VOYGR-12 plant could be worth more than $6 billion, though these estimates are based off older figures which included surging raw material costs post-pandemic. A rough estimate based on these figures places NuScale’s 1.85 GW project with Standard Power at around $12.4 billion. These estimates also does not include future revenue opportunities related to plant construction, such as fuel supply, refueling services, parts procurement, training, testing and more.

A Note into Rising Costs – UAMPS and NuScale’s Terminated Agreement

It’s vital to touch on the revenue opportunity per GW here, as NuScale’s estimated module pricing stems from its previous engagement with UAMPS, which was expected to be the first commercial SMR in the US.

The two terminated the project in late 2023 due to a significant 120% increase in expected total costs. The project was initially estimated to cost $3 billion in 2015 for a 12-module 600 MWe plant, though this was later increased to $4.2 billion as NuScale upsized to a 720 MWe plant in 2018. By 2020, estimated costs had risen to $6.1 billion, and by 2023, the project’s Class 3 estimate was placed at nearly $9.3 billion for a downsized 6-module, 462 MWe design. This included owner’s costs, interest rate assumptions, construction, operating and maintenance costs, but excluded DOE cost-sharing payments.

Looking at costs per MWh of electricity generated or cost per kW shows why UAMPS and NuScale terminated the project. It was initially proposed with a targeted power cost of $55 per MWh, but was later raised slightly to $58 per MWh by 2021, per IEEFA.

Source: IEEFA IEEFA

However, by 2023, the targeted power price surged to $119 per MWh, or $89 per MWh including a $30 IRA subsidy. This was driven by a 75% increase in construction costs, including a 54% increase in fabricated steel plates, 106% increase in carbon steel piping, 25% increase in electrical equipment and 32% increase in copper wire and cable. It also made the project far more expensive than new combined-cycle gas plants, which are estimated to have power costs around $50 per MWh, undercutting the benefits of selecting SMRs for power.

On a per kW basis, the project cost had risen nearly 5x by the time of termination, given the increased construction costs and plant downsize. At the initial 12-module, 600 MWe design, cost per kW was placed at ~$5,000, but then decreased to $4,200 to $4,300 when NuScale upsized to the 720 MWe plant design. By 2023, the downsized design and increased costs had pushed cost per kW up 5x to ~$20,140.

The failed project demonstrated that SMR costs likely “will be much higher than has been acknowledged, and the prices of the power produced by those SMRs will be much more expensive,” which could weigh on commercial viability, per the IEEFA.

Live Build with RoPower Project and Customer Update

While the UAMPS termination was a bit of a blow to NuScale’s progress, it still has one live potential project with Romania’s RoPower. NuScale is working as a subcontractor to major stakeholder Fluor on its phase 2 front-end engineering and design (FEED) services contract for RoPower’s Doicesti SMR plant.

Fluor signed the phase 2 FEED contract in June 2024 with work expected to be completed by Q4 2025, with a possible extension to Q1 2026. NuScale’s primary deliverable as a subcontractor to Fluor is the Class 3 cost estimate by this fall. NuScale said it is in discussions to extend the project into the detailed design phase, which would enable a final investment decision application to be submitted to the Romanian government by Q1 or Q2 2026. The project is expected to use a six-module plant design for 462 MWe power output, with deployment expected to occur by 2030, assuming the project moves ahead.

RoPower is NuScale’s source of revenue and cash flow at the moment, driving more than $46 million in revenue over the last two quarters. Revenue for NuScale is expected to be flat to slightly up from Q1’s $13.2 million through the remainder of the contract’s duration.

NuScale also has an agreement with Standard Power to deliver up to 24 of its 77MWe modules (~1.85 GW) for future data center sites in Ohio and Pennsylvania, with Standard Power expecting operations to commence as soon as 2029, though this may be an aggressive timeline.

NuScale Aiming to have First Firm Order in 2025

Although it is progressing with RoPower ahead of the final investment decision, NuScale is still working to record its first firm, binding customer order, with management confident in achieving this sometime in 2025.

It’s critical that NuScale sign a firm deal this year to ensure it reaches commercialization by 2030, since it has approximately a five-year timeline from contract signing to commercial deployment. Should it not record a firm order this year, commercialization will likely be pushed towards 2031.

While it continues to progress on a firm order, management shared that this customer is likely to be a power plant operator, hinting that the offtaker (actual end customer buying the power) would probably be a tier 1 hyperscaler or AI developer:

“We talk about who the customer is and the possibility that one of our customers is a hyperscaler Tier 1 data center AI developer. And I think it's important to understand that our projects are complex. And really, the customer, in many cases, will be the developer who's developing the plant and purchasing the SMRs from NuScale. Supporting that idea will be a power purchase agreement and a buyer and a power user, which we anticipate would be a Tier 1 data center or AI developer.”

Yet, the lack of a firm customer order has led some analysts to shift to a more hesitant stance. NuScale’s shares declined double-digits as BTIG downgraded shares on June 25, stating that they are waiting “for the backlog to materialize and to get more clarity around project economics.”

ENTRA1 Partnership for Commercialization, Doosan for Manufacturing

NuScale is tapping different partners to aid in manufacturing and commercialization, signing an exclusive deal with ENTRA1 to help commercialize its modules and with South Korea’s Doosan for production.

Under its partnership with ENTRA1, NuScale will simply provide its SMR modules to ENTRA1 while ENTRA1 will develop, finance, own and operate power plants globally. This frees NuScale up from capital-intensive plant development, but also limits its involvement post-deployment. The two say that this structure offers flexibility for customers, where they can either assume full transfer of ownership of the plants from ENTRA1, serve as an operator, or simply buy the power under long-term power purchase agreements.

On the manufacturing side, NuScale is tapping Doosan for primary module production and BWXT Canada for module fabrication services. Doosan, which had invested a total of $104 million NuScale through 2019 to 2021, will produce and supply core SMR components for the RoPower and Standard Power projects, and likely any subsequent projects. Doosan began manufacturing components in 2023 for the now-terminated UAMPs project.

NuScale noted that it recently placed an order for six more modules to bring its total in production with Doosan up to 12, supporting approximately 1 GW of projects. Analysts asked about manufacturing capacity given that it is outsourced, wanting to know if NuScale could support concurrent larger-scale deployments by the early 2030s, such as two 12-module deployments.

Management explained that Doosan has capacity to produce around 20 modules annually, adding that they are unsure if they will be able to fulfill multiple orders at once in the first year. CFO Robert Hamady said that NuScale will put more money into the supply chain to boost capacity once it has a firm contract, and that the startup will be “in a great place if our biggest challenge is keeping up with orders.” Looking at Oklo’s backlog of 14.1 GW requiring 160+ modules, NuScale’s backlog in GW is likely to remain limited in scope and size given that Doosan’s annual capacity cannot support its 1.85 GW deal with Standard Power for 24 modules.

Diversified Revenue Opportunities

Unlike Oklo’s build, own and operate approach, NuScale’s approach provides it with longer-term revenue and cash flow streams, that commence as early as five years before commercial operations begin.

Below is a diagram of different revenue streams and timing of recognition, up until commercial operations commence for the module and afterwards.

Source: NuScale NuScale

In the first phase prior to deployment, NuScale’s initial revenue streams will stem from licensing and support, such as its current scope of work with RoPower. This then transitions into training, inspection and testing leading up to two years prior to the module deployment. At and after deployment, NuScale’s revenue opportunities then shift to auxiliary needs including fuel supply, engineering management, parts and procurement and more.

Management explained on Q1’s call that this approach will provide them with immediate revenue and cash flow upon signing a firm contract:

“We believe that once we sign a contract with a major customer, we will receive payments in relation to the modules. We'll receive positive cash flow. … In relation to the sale of the modules, we expect somewhere around 25% of cost of modules to come in the first year and for NuScale to be cash flow positive from that perspective. [A firm contract] will be pivotal in terms of pushing our balance sheet and our cash flow towards a cash flow positive position.”

This is why NuScale is emphasizing a firm contract and customer order by the end of 2025, and why they doubled modules in production to 12 despite having an order. Signing a firm contract will bring immediate revenue and cash flow to slow cash burn, while simultaneously providing a much larger long-term revenue stream.

First to Receive NRC Approval & Regulatory Update

In the SMR field, NuScale believes it holds a significant first-mover advantage as the first and only firm to have received standard design certification from the NRC for its modules. NuScale’s modules are powered by proven light-water reactor tech, essentially a micro-sized version of a conventional large-scale reactor.

NuScale recently received NRC approval for its 77 MWe US460 design, which was based in part off its 2023 NRC certified US600 50MWe design. The approval means that the 77 MWe design meets rigorous safety standards and is now approved for use in the US without further review, valid for 15 years. The design can also be referenced by developers and other companies in COLA or construction permits.

Despite not commenting much on the regulatory side in Q1, management made an effort to point out how much of a first mover advantage they believe they hold after the recent second NRC approval. CEO John Hopkins said that competing LWR reactors are “still in what we call the demonstration phase, which means they still require a minimum of 4 years of operation before securing U.S. Nuclear Regulatory approval for commercial deployment.”

However, even with NuScale’s NRC approval, customers still have to proceed with site-specific COLAs for each plant they plan to build, which lengthens deployment timelines with the extra regulatory steps.

Financials – Revenue Ramp Expected in 2026, 2027, Cash Runway Quickly Improved

Though NuScale’s first twelve SMR modules are in production, it does not anticipate commercial deployment with its first project and only customer, RoPower, until 2030. Despite this, revenue is expected to begin ramping through 2026 and 2027 due to NuScale’s multiple revenue streams prior to deployment.

Revenue Ramp Projected for FY26

Revenue is currently tied to consulting and front-end engineering and design (FEED) services for NuScale and Fluor’s partnership with RoPower. NuScale reported $34.2 million in revenue in Q4 and $13.4 million in Q1. Q4’s print was notably higher due to one-time tech licensing payments received in conjunction with RoPower.

Through Q4, NuScale is expected to see revenue hover in the low-teens each quarter, ranging from $11.9 million in Q2 to $14 million in Q4.

For FY25, NuScale’s revenue is estimated to be $50.4 million, for 36.1% YoY growth, with the low growth rate being impacted by Q4’s $34.2 million print. Beyond FY25, NuScale is expected to see revenue begin to ramp rapidly. FY26 revenue is expected to increase 226% YoY to $164.4 million before slowing to 120% YoY to $361.8 million in FY27.

Margins

Margins are quite lumpy given the scope of work under the RoPower project, with gross margin fluctuating quite significantly. It’s also not a clear sign of the future business once SMRs are deployed and ramp.

Gross margin had soared to 91.1% in Q4 on the substantial increase in revenue related to the RoPower projected, but has since dropped to 52.4% in Q1. Gross margin had been as low as 12.1% in Q2 2024.

Operating margin remains deeply red considering revenue streams are still minimal. Operating margin was (34.6%) in Q4, but reversed back to (261.4%) as of Q1. With operating expenses hovering in the low-$40 million range with revenue in the low-teens, operating margin is likely to remain >(200%) through FY25.

However, NuScale is realizing meaningful cost reductions, noting in Q1 that it has lowered its average quarterly operating expenses by ~$28 million, from $69.9 million in 2023 to $42.1 million on a TTM basis. NuScale says this is generating annualized savings of $111.2 million.

This improvement is visible when looking at operating losses on an annual basis. FY23 operating loss was ($275.6) million, though NuScale cut that in half in FY24 to just ($138.7) million.

EPS

NuScale is not expected to break even until 2030, with losses accumulating until then. Quarterly adjusted EPS was ($0.11) in Q1 and is expected to remain flat at that level for at least the next four quarters. Q4’s adjusted EPS of ($0.77) was negatively impacted by the $227.7 million redemption of its warrants.

Cash Flows, Cash Burn and Liquidity

The Q4 warrant redemption significantly bolstered NuScale’s balance sheet, helping propel it to more than half a billion in cash and equivalents. This combines with improvements in cash burn to provide the company with a much longer cash runway.

As a result of its increasing balance sheet and decreasing cash burn rate over the last few quarters, NuScale has substantially lengthened its cash runway. As of Q1, NuScale had $521.4 million in cash and equivalents on its balance sheet. This major improvement from $156.6 million in Q3 (less than its annual burn rate) was due to the warrant redemption and 4.5 million ATM offering, which provided $227.7 million and $102.4 million in proceeds respectively.

In terms of cash runway, these actions the last two quarters increased NuScale’s runway by nearly 18 quarters. As of Q2, NuScale’s cash runway was less than one year, but now as of Q1, NuScale has extended this runway to more than 21 quarters. The company said in Q1 that the liquidity improvement was due to financing activities as well as reduced operating expenses and payments from the FEED 2 contract.

CFO Robert Hamady offered some color on liquidity and cash burn in Q1’s analyst Q&A:

“So if I look at just like an OpEx cash burn, it's still in that range, let's say, $40 million to $45 million a quarter. We'll probably push it up a little bit as we invest in our supply chain. That's something that we're actively doing now because we anticipate a near-term project, and we're focused on supply chain manufacturing and delivery dates. …

We think we have 2 years or so of operating runway, just based on where we are today, just based on what cash we have and what we bring in, what we spend. And that will change only when we get a project, and I think it will change in a positive way for the company and for our liquidity position.”

He further clarified later that “the $40 million to $45 million vis-a-vis $500-plus million worth of cash would seem to instigate an idea of greater than 2 years' runway, especially when you add some revenue there, right, because it lowers the burn rate. But I'd like to keep 2 years runway.”

This comment implies that NuScale has a longer cash runway, if it does dwindle to the point that this runway encroaches on two years or shrinks to less than two years, capital raises are likely in the picture.

Customer Concentration Risk

NuScale is exposed to significant customer concentration risk, with RoPower its only current customer and primary source of revenue. NuScale says that it has “robust business development activity including advancements with prospective data center/artificial intelligence (AI) customers,” though it has received no firm orders yet.

Should RoPower determine not to proceed with the plant, NuScale could witness substantial fallout given that this is its first commercial project and proof of SMR market fit. This could translate into several billions of lost revenue along with more questions about a path to break-even and commercial viability of its modules.

NuScale also signed an agreement with Standard Power in 2023 to provide 24 of its 77 MWe reactors, or up to 1.85 GW, to power two future data center sites in Pennsylvania and Ohio. Standard Power expects to be operational by 2029, and the 77 MWe design just received NRC approval at the end of May 2025, paving the way for this project to proceed.

Valuation

Similar to Oklo, the timing of NuScale’s deployments and revenue ramp open the door to some execution risk. However, it is valued at a nearly 40% discount to Oklo at $4.8 billion with revenue expected to ramp much sooner.

Based on current forecasts, NuScale is valued at approximately 29.3x FY26 revenue with 226% YoY growth expected, and then 13.3x FY27 revenue with triple-digit growth again.

On a more direct comparison to Oklo on FY30 revenue, NuScale trades at just a fraction of its rival, at 3.5x expected revenue of $1.13 billion versus Oklo at 30x revenue of $257 million. This gap implies a significant time-to-market premium for Oklo, given NuScale is expected to see revenue ramp years earlier.

Conclusion

Despite its first-mover advantage with NRC approval and its first modules in production, NuScale lacks the one piece needed for its commercialization pathway to materialize – a firm customer order. These orders are more likely to come in as Big Tech scales towards Blackwell Ultra or Nvidia’s upcoming Rubin generation due to the sheer increase in power in these rack-level systems.

NuScale is expected to see revenue ramp rather quickly over the next two years with triple-digit growth projected, though capacity limits revenue upside at just 20 modules per year. As is the case with Oklo, this is a high-risk, speculative play on nuclear adoption for data center power demand. Any slight delays in NuScale’s commercialization or ramp given its limited capacity and lack of a firm order yet could significantly push back this growth curve, and thus delay profitability and cash flow growth.

Damien Robbins, Equity Analyst at I/O Fund contributed to this analysis.

Every Thursday at 4:30 pm Eastern, the I/O Fund team holds a webinar for premium members to discuss how to navigate the broad market, as well as various stock and crypto entries and exits. Beth Kindig offers weekly deep dives including lesser-known cryptocurrencies and AI stocks, plus the team offers trade alerts. The I/O Fund team is one of the only audited portfolios available to individual investors. If you’d like to subscribe to the Advanced Market Signals plan, email us at premium@io-fund.com4:30 pm Eastern, the I/O Fund team holds a webinar for premium members to discuss how to navigate the broad market, as well as various stock and crypto entries and exits. Beth Kindig offers weekly deep dives including lesser-known cryptocurrencies and AI stocks, plus the team offers trade alerts. The I/O Fund team is one of the only audited portfolios available to individual investors. If you’d like to subscribe to the Advanced Market Signals plan, email us at premium@io-fund.compremium@io-fund.com.

Recommended Reading:

Oracle Cloud May Grow Much Faster than Big 3

Posted on July 18, 2025June 30, 2026 by io-fund

This article is a continuation of our free newsletter from July 17, Can Oracle Become the Next $1 Trillion AI Stock?

Find out the following below:

If Oracle can become the next $1 trillion AI stock – or if Stargate is a one-time deal that will leave the stock flat from exorbitant GPU costs
Bull, base and bear case scenarios, with detailed revenue and segment projections showing where and when revenue upside lies
An invitation to our next webinar – exclusive only to subscribers. In our upcoming 1-hour webinar, we will detail our buy, trim and sell zones for Oracle and other leading AI stocks.

Oracle Cloud May Grow Much Faster than Big 3

Though Oracle is growing off a much smaller cloud base than say Azure, robust IaaS momentum could drive its Cloud growth at a much faster rate than the Big 3 – defined as Microsoft, Amazon and Alphabet — over the next few years.

This rapid IaaS growth could fuel a 40% CAGR for Oracle’s total Cloud growth by FY28, taking its Cloud segment from $24.4 billion to $66 billion. This 40% CAGR will far outpace AWS’ growth in the high-teens, and Google Cloud and Azure in the high-20% to low-30% range.

While this represents a significant reshapement of Oracle’s business model towards high-growth cloud, it also implies that Oracle is aiming to be much more competitive in AI and firmly establish itself as the fourth AI hyperscaler. Despite coming off a much smaller revenue base, Oracle’s forecasted 40% cloud CAGR suggests that it will start to take market share from the Big 3.

Scenario Analysis

Quick Look into FY26, Long-term Targets

Before we present the scenario analysis, it’s prudent to briefly touch upon Oracle’s FY26 guidance and long-term targets, as these pertain directly to the three scenarios.

For FY26, Oracle slightly raised its revenue guidance, now expecting revenue to be $67 billion, for YoY growth of 16.7%, up $1 billion from its prior view for $66 billion. Management remains confident in exceeding its 20% growth target for FY27, or $80.4 billion.

Oracle has maintained its FY29 revenue target of $104 billion, which would require growth at a 14.5% CAGR since FY24. Given FY26 was raised, if the company maintains that CAGR then it will see 4% higher revenue in FY29 versus that target. For comparison, the consensus estimate for FY29 has risen $6 billion over the last three months, from $100.5 billion to $106.5 billion.

Bull Case Projects 55% Upside

Based on the model below through FY29, Oracle could see IaaS maintain a robust 59% 4-year CAGR to nearly $65 billion in revenue, with a deceleration to 40% YoY in FY29. This could drive Cloud revenue at nearly a 39% CAGR to more than $90 billion, or 85% of revenue, doubling its share from 43% in FY25.

This also assumes an (8%) CAGR in licensing, support, hardware and other revenue, driven by increased cannibalization of on-prem licensing beyond FY27 as Oracle’s cloud shift accelerates.

Overall, this bull case projects revenue slightly below FY27’s target, but sees much greater upside versus consensus by FY29 as the $30 billion deal begins to kick in and as Oracle benefits from increasing synergies with OpenAI.

To note, OpenAI is optimistically forecasting $125 billion in revenue by 2029, which could translate into tens of billions to Oracle from AI infrastructure costs. This forecast also likely hinges on limited hardware constraints to allow Oracle to meet such high demand.

However, high capex requirements to build out multiple GWs of capacity are likely to keep free cash flow limited – this model projects FCF margins in the mid-single digits by FY29. On a cumulative basis from FY26 through FY29, free cash flow is expected to under $7 billion.

The bull case scenario also assumes adjusted operating margins are slightly softer in FY26 and FY27 as amortization of intangible assets decreases, before expanding into FY29 as cloud and IaaS mix expands substantially.

Adjusted net margin is expected to increase slightly towards 32%, on the possibility that high capex requirements lead to debt raises and increased interest expenses.

Under this scenario, topline growth is expected to remain rather robust at an 18.8% CAGR from FY25, more than four points ahead of Oracle’s guided 14.5% CAGR. Peak revenue growth is forecast for FY28 at almost 22%, while EPS growth is expected to be >20% YoY for FY27 through FY29. As a result, Oracle is assigned a 10x P/S multiple and a 31x P/E multiple.

This is approximately in line with Oracle’s peak top-line valuations over the past five years and a 15% discount on the bottom line, accounting for a slight YoY deceleration in FY29. However, this is around a 40% premium to Oracle’s five-year average P/S multiple and a 30% premium to its five-year average P/E.

This valuation is also approximately in line with Microsoft’s current five-year average forward P/S and P/E multiples of 10.7x and 30.5x. To note, Oracle is projected to have a much higher cloud mix at 85% of revenue, growing at a 39% CAGR, versus Microsoft’s Intelligent Cloud at 66% share by 2029 on a 16% CAGR.

These assumptions return a price target of $386, or a 55% return, propelling Oracle into the $1 trillion club.

Base Case Sees Potential 27% Upside

The base case scenario assumes revenue remains roughly in line with consensus estimates through FY27, with low-single digit upside in FY28 and FY29 as Oracle begins to recognize growth from the recent mega-deal.

Under this scenario, IaaS growth will decelerate a bit quicker into FY29, with revenue at 36% YoY versus 40% YoY in the bull case scenario. This still results in a 57% IaaS CAGR to $62.3 billion in revenue, driving a 37% cloud revenue CAGR, nearly 2 points below the bull case.

In the base case, Oracle is assumed to maintain operating margin in the 43% level moving forward, with some slight net margin expansion to the 31% range. This would project FY29 EPS of $11.69, or almost 2% ahead of consensus.

This scenario sees Oracle maintaining solid topline and bottom-line growth rates, at almost 16% YoY for revenue and 18% YoY for adjusted EPS in FY29. In this, Oracle is assigned an 8.5x P/S multiple and a 27.1x P/E multiple, a 15% premium to its current five-year averages of 6.9x and 23.7x, respectively, as the company exceeds its long-term targets on strong cloud momentum.

These assumptions return a $316 price target, or 27% upside from current levels.

Bear Case Projects Flat Return

The bear case scenario assumes revenue again remains roughly in line with consensus estimates through FY27, before lagging consensus in FY28 and FY29. However, this scenario still assumes Oracle remains ahead of its FY29 target due to the mega-deal, as even a delayed ramp to 2029 still likely derisks that $104 billion revenue target.

IaaS revenue is projected at a 56% CAGR through FY29, with growth decelerating rather sharply from 70% in FY27 to 34% by FY29. SaaS growth is estimated at a slower 11% CAGR, with mid-single digit growth in FY29. This combines for a 35% cloud revenue CAGR, nearly 4 points slower than the bull case and 2 points slower than the base case. Even with these assumptions, FY29 revenue is still projected more than 1% ahead of Oracle’s target at $105.2 billion.

In this scenario, margins are projected to decline marginally through FY27 on account of that decreasing amortization expense, with further softness into FY29 from an aggressive pursuit of capacity expansion and high capex in an effort to support stronger IaaS growth.

This also assumes that Oracle will turn to the debt markets to fuel this capacity expansion, adding hundreds of millions to interest expenses and creating up to a ($0.20) EPS headwind by FY29.

As such, topline growth is projected to peak just below 20% in FY27, before decelerating to the low 14% range by FY29, resulting in a four-year CAGR of 16.8%. Similarly, EPS growth would peak at 21% in FY27 before decelerating 9 points to 12% YoY by FY29.

As such, Oracle is assigned a 7x P/S multiple and 23.3x P/E multiple, approximately in line with its current five-year averages, accounting for this two-year deceleration in growth rates. It’s also possible under this scenario that cumulative FCF generation through FY29 remains negative to barely positive, assuming capex growth comes in above >15% YoY in both FY28 and FY29.

This scenario would return a price target of $249, or essentially flat with its current share price.

Conclusion

Oracle laid out rapid growth targets for fiscal 2026 driven by strong AI momentum, forecasting cloud infrastructure growth to accelerate 20 points and total cloud revenue to accelerate 16 points. Coupled with r >100% RPO growth guidance, its expanded arrangement with OpenAI and its mega $30 billion/year cloud deal, confidence is increasing in Oracle’s long term cloud trajectory.

Join our next 1-hour webinar held Thursdays at 4:30 p.m. Eastern to discuss buy, sell, trim levels for Oracles and other leading AI stocks. To receive $100 off our Advanced tier, use code ADVANCED100 or click hereclick here and email your request to upgrade.

Recommended Reading:

Can Oracle Become the Next $1 Trillion AI Stock?

Posted on July 18, 2025June 30, 2026 by io-fund

Oracle is a stock that was long-forgotten about in the cloud era when the Big 3 — Microsoft, Google and Amazon — and many best-of-breed companies stole the spotlight. Oracle struggled to report growth of any kind with some years in the red and other years rarely reporting over 5% growth during cloud’s glory years.

But today, Oracle wants to be taken seriously as a strong AI contender, and it is quickly positioning itself to lead among Microsoft, Amazon and Alphabet when it comes to AI-driven cloud growth over the next few years.

The stock has done quite well off the April lows following some impressive growth forecasts in June, including guidance for a 20-point acceleration in Cloud IaaS growth to >70% YoY. Yet it will take a lot more than one earnings report to convince the market that the enterprise giant has staying power.

Below, we look into the sleepy enterprise giant to help determine if Oracle has what it takes to compete with innovative best-of-breed companies and the Big 3 who are spending tens of billions to fortify their moats in the fierce arms race of AI.

The Economics of OpenAI Stargate Expansion

At the beginning of July, OpenAI signed an agreement to rent 4.5 GW of data center capacity from Oracle in the US. This marks one of the largest single data center leasing deals in history, with 4.5 GW being approximately 25% of the current operational data center capacity in the US.

Based on prior estimates for its Stargate data center in Abilene, the 4.5 GW expansion could require as many as 2.25 million GPUs, or ~$94 billion worth of Nvidia’s GB200 NVL72 racks.

Analysts from BNP Paribas have estimated that 4.5 GW of capacity could generate $30 billion to $60 billion in annual revenue for Oracle, depending on hourly GPU rental prices.

However, 4.5 GW of capacity is not cheap. BNP places the cost of this buildout at $180 billion to $225 billion, stating that it is unclear if Oracle will have to finance the project itself or if Stargate partners including Microsoft and SoftBank will contribute.

Oracle had reportedly purchased ~400,000 Blackwell GPUs worth $40 billion in late May to expand the Abilene data center to support OpenAI, yet this is likely just a fraction of what is needed for this new mega-deal.

Oracle to See $30 Billion In Annual Revenue from Major Cloud Deal

Just a few days prior to the 4.5 GW announcement, Oracle announced that it had signed multiple cloud service agreements, including a major contract expected to contribute more than $30 billion in annual revenue starting in FY28. To put that in perspective, this is nearly 25% larger than Oracle’s entire cloud business in FY25, and 3x its entire cloud infrastructure business.

Jefferies analysts pointed out that Street estimates currently model $46 billion in Cloud IaaS revenue in fiscal 2028 on $93 billion total revenue, suggesting the new deal alone is worth two-thirds of projected IaaS revenue. Jefferies added that while the ramp is years out, the magnitude unlocks potential upward revisions to Oracle’s fiscal 2029 revenue target of $104 billion.

Oracle did not name the customer, though it is widely expected that OpenAI is behind this lucrative deal. This would place each GW of capacity at around $6.67 billion in revenue annually, or ~$555 million per month, based on the 4.5 GW scale.

Oracle’s Cloud IaaS Guided to 70% Growth this Fiscal Year

During its Q4 report in early June, Oracle projected substantial acceleration in its cloud business in fiscal 2026, fueled by strong AI demand and cloud consumption.

Total cloud growth (IaaS & SaaS) was guided to accelerate to more than 40% in FY26 up from 24% in FY25. This will equal $34.2 billion in revenue and implies the segment may exit FY26 well in the 50% YoY range.
Cloud IaaS (OCI) growth was guided to accelerate to more than 70% in FY26 up from 50% in FY25, equaling $17.5 billion in revenue. Assuming sequential revenue growth is stronger in the back half, IaaS revenue could potentially accelerate to more than 30 points to the low to mid-80% range.
Oracle Cloud Infrastructure consumption revenue is expected to grow faster than the 62% YoY increase reported in Q4.

Oracle’s Cloud revenue is showing signs of accelerating with Q4 revenue rising 27% YoY.

Below is a chart showing how swiftly Oracle is expecting to accelerate IaaS revenue through FY26. Actual growth rates may vary should Oracle have a stronger H1 than H2.

In one possible scenario, Oracle could see Cloud Infrastructure revenue accelerating more than 30 points to the mid-80% range in FY 2026.

Strong Database Growth Supports Oracle’s Cloud Acceleration

In our free stock analysis two months ago, we stated that Oracle’s ability to deliver low latency, high-performance AI at scale gives it a slight edge in the cloud. Oracle’s vector capabilities stand out given its database roots, and Oracle expects cloud database to be its third revenue driver in FY26 (behind IaaS and strategic SaaS). This is notable considering the bulk of Oracle’s database presence remains on-prem.

While cloud database growth outpaces SaaS growth at 31% YoY, it’s coming off quite a small base, at just $2.6 billion annualized in Q4, or ~$650 million quarterly. Within this segment, there were pockets of rapid growth, with Q4 MultiCloud database revenue from Azure, AWS and Google Cloud increasing 115% QoQ.

For FY26, Oracle projected triple-digit growth in MultiCloud database revenue to continue, building on Q4’s strong momentum, as it signaled an intent to double and triple cloud-dedicated datacenters. Oracle has 23 MultiCloud data centers in operation currently, and it is working to triple capacity, building 47 more MultiCloud data centers over the next 12 months.

Additionally, Oracle is aiming to double its Oracle Cloud@Customer data center footprint, from 29 live datacenters with 30 more being built in FY26. Management noted that Cloud@Customer data center revenue increased 104% YoY, and doubling capacity paves the way for this robust momentum to drive strong growth in FY26.

Considering its high on-prem presence, analysts questioned management about cloud database’s contribution to the 70% IaaS growth target and opportunities from cloud migration. Management did not place an exact number on FY26’s contribution, though CTO and chair Larry Ellison implied that every $1 billion in database support revenue could be worth $5 billion in the cloud, due to compute, networking, storage and other fees associated with cloud databases.

For another look at Oracle’s AI advantages and vector capabilities, read Oracle Stock Outlook: Revenue Could Double by FY2029, yet Targets Seem Lofty.Oracle Stock Outlook: Revenue Could Double by FY2029, yet Targets Seem Lofty.

Oracle Expects RPO to Double to More Than a Quarter Trillion

One of the more impressive forecasts Oracle doled out in Q4 was its RPO growth target for FY26. Management stated that RPO was expected to increase more than 100% in the upcoming fiscal year, which would place RPO at well over one-quarter trillion.

Oracle reported $138 billion in RPO in Q4, up $8 billion QoQ and up 41% YoY. Thus, management is forecasting RPO to rise to approximately $275 billion to $280 billion, a significant YoY increase given Oracle only added $40 billion YoY in RPO in FY25. Oracle has previously been vocal about how its cost-efficient AI services with multi-cloud database support help it win larger deals, and this is likely a major factor in FY26’s lofty RPO target.

A simple estimate of the trajectory of RPO growth Oracle is guiding for FY26.

Cloud RPO has been robust and will likely be primary driver in reaching this guidance, boosted by Oracle’s involvement in Stargate. Cloud RPO is approximately $110 billion, rising 56% YoY in FY25, on top of 80% YoY growth in FY24. This growth outpaces both Amazon and Microsoft, with Amazon’s backlog up 20% YoY to $189 billion, while Microsoft’s rose 33% YoY to $315 billion.

Analysts had questioned management about Stargate’s impact on RPO and IaaS growth forecasts, and while they were quite vague about the project’s contribution, CTO and chairman Larry Ellison said that “if Stargate turns out to be everything is advertised, then we've understated our RPO growth.” This implies further upside beyond one-quarter trillion should Stargate continue to expand towards its half a trillion dollar investment size.

Additionally, Oracle’s RPO provides strong visibility into FY26 revenue, with 33% of total RPO expected to be recognized over the next 12 months. This suggests Oracle will recognize ~$45.5 billion of its backlog, or 68% of its guided $67 billion in revenue for the year.

Understanding Oracle’s Capex Strategy Amidst Surging Demand

Capex will remain elevated in FY26 as Oracle works to build out capacity to meet high demand, especially considering management expects RPO to increase by more than $130 billion during the year. Costs for the upcoming 4.5 GW expansion are likely to be tens to hundreds of billions as well, depending on how much Oracle will contribute itself.

For FY26, Oracle expects capex to be greater than $25 billion, or up ~18% YoY from $21.2 billion in FY25. It’s also important to point out here that the $21.2 billion figure was well ahead of Oracle’s guidance from fiscal Q3 for around $16 billion in capex. This is because Q4 capex surged 55% QoQ to $9.1 billion.

CEO Safra Catz also stated on Q4’s call that she believes the $25 billion forecast “may turn out to be understated.” Think of it this way – how does Oracle bring supply to meet demand given one quarter trillion in RPO and the $30 billion deal would be more than 10x the current size of its cloud business today? Oracle is already short of capacity with cloud growth accelerating, suggesting capex may indeed need to be much higher to support these massive newer deals.

While this may seem like peanuts in the hyperscale space, with Amazon projected to spend >$105 billion this year on capex, Oracle’s spending is rather substantial relative to its size. For example, Oracle went from spending the least capex to revenue out of Big Tech in early 2022 at just 9%, yet it now committed 37% of revenue to capex in FY25. This far outpaced Meta at under 26% and Alphabet and Amazon in the mid-teens.

Oracle now spends 37% of revenue on capex, the highest among the hyperscalers, after it spent the least on capex just four years ago. Source: YChartsYCharts

Implications of Oracle’s Elevated Capex

One downside to this high relative capex is the pressure that it places on Oracle’s cash flows, especially if capex is much higher than originally anticipated. Analysts also pointed out another problematic scenario for investors that may arise from this soaring capex.

Q4’s 55% QoQ surge in capex far outpaced operating cash flow, which rose less than 4% QoQ to nearly $6.2 billion. As a result of capex being 50% higher than OCF, free cash flow was ($2.9 billion) for the quarter.

And for the first time, Oracle’s spent more on capex than it generated from operations, pushing FCF into negative territory for the year at ($394 million).

For the first time, Oracle spent more on capex than operating cash flow. Source: YChartsYCharts

For FY26, Oracle’s capex forecast risks another year where negative FCF is possible. To avoid this, Oracle would have to grow its operating cash flow in excess of 25% YoY, a significant acceleration from just 11.5% YoY growth in FY25. This raises the question of how Oracle will continue to fund dividends and buybacks while simultaneously (and aggressively) expanding capacity.

Consider Oracle’s liquidity profile as of Q4, with 9x debt to cash — $10.2 billion in cash and equivalents and $92.6 billion in debt. Maintaining this capex trajectory, potentially at $55B+ over the next two years, will likely require Oracle to raise more debt.

Analysts from Barclays already see potential for substantial debt raises through FY27 to support this capex intensity, buybacks and dividends. Barclays estimates Oracle could raise $10 billion in both FY26 and FY27 at a 5% borrowing cost, which could add up to $300 million in interest expense, weighing on earnings growth.

The payoff for this elevated capex is maintaining cloud revenue acceleration over the next eight to twelve quarters, as cloud will be instrumental in Oracle meeting or exceeding its long-term growth targets.

Can Oracle Become the Next $1 Trillion AI Stock?

The biggest question here for long-term shareholders is if this AI-driven cloud acceleration can propel Oracle to become the next $1 trillion company.

Oracle is in the middle of a profound reinvention of its business model, migrating from a low-growth, license-based model to a high-growth, AI cloud infrastructure engine. Driving this pivot are record-breaking partnerships and cloud deals and robust industry-wide momentum. Yet does the company have what it takes to succeed against a formidable trio of Microsoft, Amazon and Alphabet in the cloud?

Subscribe Below for the Following:

If Oracle can become the next $1 trillion AI stock – or if Stargate is a one-time deal that will leave the stock flat from exorbitant GPU costs
Bull, base and bear case scenarios, with detailed revenue and segment projections showing where and when revenue upside lies
An invitation to our next webinar – exclusive only to subscribers. In our upcoming 1- hour webinar, we will detail our buy, trim and sell zones for Oracle and other leading AI stocks. Subscribe to continue reading

Paid subscribers, read the complete analysisread the complete analysis.

Not ready to subscribe but want more thoughtful analysis from a top-performing team in tech? Every week, we publish free research. Sign up to our free newsletterSign up to our free newsletter

Recommended Reading: