Dell raised AI server shipments for the fiscal year from $15 billion to $20 billion, for a raise of 33%. AI servers are expected to account for nearly 19% of Dell’s revenue in FY26 following ISG growing an impressive 63% QoQ.
Management stated it was the “single largest number of customers that we sold in a quarter” and “it is the most revenue we generated to enterprise customers in a quarter to date.”
The size of AI revenue is impressive and ranks Dell in the top 10 companies on the public markets by size of AI revenue. Consider that Broadcom will deliver a similar number this year.
However, the key difference with Dell is the margins are slim – with one analyst asserting that AI server margins could be as low as 2.5% calculated off the guide. The GAAP numbers show gross margin dipped although operating margin held steady. The CFO did state margins would improve into the second half.
There were some notable comments around the enterprise AI market heating up, plus a mention of having 6700 unique customers in their pipeline (sounds like a large number when you think of the high customer concentration most hardware companies must contend with).
Overall, the margins overshadowed the otherwise exceptional AI growth. Until the margins improve, the market will not be rewarding the stock the same as its AI peers.
Revenue Growth Accelerates to 19% YoY
Dell reported $29.78 billion in revenue in Q2, coming in above the high-end of its guided $28.5-29.5 billion range and beating estimates of $29.2 billion by 2.7%. This corresponded to growth of 19.1% YoY, a sharp 14 point sequential acceleration, driven primarily by outperformance in AI servers with record shipments.
This growth is expected to be somewhat of a one-off, with Dell guiding for just 11% YoY growth in Q3 to $26.5-27.5 billion in revenue.
For FY26, Dell boosted its revenue outlook by $4 billion at midpoint, now seeing $105-109 billion, up from $101-105 billion previously; this corresponds to growth of 11.9% YoY. This places Q4 revenue at a tentative $26.8 billion, or $0.8 billion ahead of estimates.
On closer view, Dell’s FY26 guidance raise was $1 billion lower than its AI server guidance raise of ~$5 billion, suggesting that there may be some demand-related headwinds in PCs or other segments it needs to parse through.
Key Segments
Infrastructure Solutions Group (ISG) up 63% QoQ, Server Shipments up 356% QoQ
As expected, Dell’s ISG revenue experienced significant growth in Q2, up 44.2% YoY and 62.8% QoQ, driven by strong outperformance in AI server shipments in the quarter. For Q3, Dell guided for ISG growth in the low-20% range YoY, likely driven by normalizing AI server shipments.
Q2’s AI server shipments came in well above Dell’s $7 billion guide, rising 356% QoQ and 165% YoY to $8.2 billion. Although orders more than halved sequentially to $5.6 billion, Dell’s AI server backlog remained elevated at $11.7 billion.
As a result of the strong shipments in Q2, taking 1H to $10 billion, Dell raised its AI server shipment outlook for the year from $15 billion to $20 billion. While this is a welcome increase, the dichotomy between shipments and orders raises some concern, with Dell shipping only $10 billion of its $18 billion in orders booked so far this fiscal year.
Additionally, the FY guide again points to normalizing AI server shipments of ~$5 billion/quarter in 2H, a slight uptick from our Q1 readthrough of $3.5-4 billion, but still a bit soft when factoring in backlog and orders.
Within ISG:
Servers and Networking revenue: $12.94 billion, up 68.7% YoY and 104.8% QoQ. This is the strongest combination of YoY and QoQ growth for the subsegment since Q2’25, driven primarily by that AI server strength.
Storage revenue: $3.86 billion, down 3.0% YoY and down 3.5% QoQ.
ISG operating incomewas$1.47 billion, up 14.5% YoY and up 47.3% QoQ. However, this reflected an operating margin of 8.8%, down from 9.7% in Q1’26 and down from the 11.0% reported in Q2’25.
The CFO stated margins would improve in H2: “Our ISG operating income rate was down year-over-year to 8.8% of revenue. As we have outlined before, the mix of our AI business will have an impact on our margin rates. In the second quarter, we saw a significant shift in our mix to AI as the team executed very well and drove record AI shipments. This was the primary driver of our operating income rate this quarter, partially offset by lower operating expenses. Given our engineering differentiation and integration, we expect our AI margin rates to improve in the second half.”
Client Solutions Group (CSG)
CSG revenue rose less than 1% YoY and was approximately flat QoQ to $12.50 billion, as Commercial momentum faded. CSG operating income was $803 million, or a 6.4% margin, up from 5.2% in Q1’26 but down from 6.6% in Q2’25.
Commercial revenue of $10.78 billion declined (2.4%) QoQ and decelerated 7 points to 2.1% YoY, despite Dell noting that commercial PC demand grew YoY for the sixth straight quarter. Consumer revenue rebounded 17.7% QoQ but remained down (7.3%) YoY to $1.72 billion.
For Q3, Dell guided for CSG revenue growth to rebound to the mid single-digits.
Gross Margins Dip Below 20%, Operating Margins Show Strength
Dell’s gross margins dipped below 20% for the first time in three years, and this margin compression was a key focus point on the earnings call, as analysts wanted to understand underlying drivers and forward expectations.
Management indicated that margin degradation was driven by AI server mix shift but expects to see improvements into the back half of FY26. We will continue to monitor how gross margins trend in future reports.
In Q2, GAAP gross margin contracted nearly 300 bps YoY and QoQ to 18.3%, while adjusted gross margin was nearly 400 bps lower YoY to 18.7%. Operating margins showed some strength and reflected economies of scale driven by top-line growth, as GAAP operating margin expanded both YoY and QoQ.
GAAP operating margin was 5.9%, up from 5.4% in the year ago quarter and 5.0% in Q1.
Adjusted operating margin was 7.7%, down slightly from 8.1% in the year ago quarter but up from 7.1% in Q1.
GAAP net margin was 3.9%, up from 3.4% in the year ago quarter but down slightly from 4.1% in Q1.
Adjusted net margin was 5.3%, down slightly from 5.5% in the year ago quarter but up from 4.6% in Q1.
EPS
Dell reported $2.32 in adjusted EPS in Q2, a marginal $0.03 beat versus estimates. Adjusted EPS growth did reaccelerate to 22.8% this quarter, though this is likely to be the peak growth quarter for the fiscal year as the margin compression led to a soft Q3 EPS guide.
For Q3, Dell guided for just 11% YoY growth to $2.45 in adjusted EPS at midpoint, below estimates for 18.5% growth to $2.55. For Q4, analysts are projecting 11% growth to $2.98.
For FY26, Dell updated its adjusted EPS guidance to $9.55 at midpoint, up 25% YoY, ahead of estimates for $9.37. This will likely force revisions in Q4 to move 7-9% higher considering adjusted EPS is pegged at approx. $6.32 through Q3.
Looking out to FY27, EPS growth is expected to cool to 15.1% YoY to $10.79.
Cash Flow Margins Decline Sequentially as Accounts Receivable Surge
While Dell did report record 1H operating cash flow at $5.3 billion, cash flow margins contracted from Q1 as accounts receivables surged more than 50% QoQ.
Cash & Cash Equivalents were $8.15 billion in Q2, up from $7.7 billion in Q1. Debt was largely flat with Q1 at $28.7B.
Operating cash flow came in at $2.54 billion in Q2, down from to $2.8 billion in Q1 but nearly doubling against $1.34B in Q2’25. OCF margin was 8.5%, down from 12.0% reported in Q1’26 but up from 5.4% reported in Q2’25.
Free cash flow came in at $1.87 billion, down from $2.23 billion in Q1 and up from $1.28 billion in Q2’25. FCF margin was 6.3%, down from 9.5% reported in Q1’26 but up from 5.1% reported in Q2’25.
Behind the sequential decline in cash flow margins was a 53.5% QoQ increase in accounts receivable to $15.02 billion. On a YoY basis, AR was up 31.9% YoY. DSO as of Q2’26 was ~46 days, compared to ~41 days in Q1 FY26 and ~39 days in Q2 FY25. This jump corresponds with the explosion in AI Server shipments, as revenue recognition occurs at shipment and cash is collected later-on. AI server customers are large enterprise & CSP buyers which may negotiate longer credit terms compared to the consumer / SMB segments. The jump in AR nearly matches the $6.4B sequential increase in payables which largely offsets the impact to cash conversion. The combination in higher AR & AP should indicate to investors that Dell may be carrying larger working capital load tied to the ramp of Blackwell – both collecting later from customers and paying later to suppliers.
Inventories of $7.21B reflects a (2.8%) decline compared to $7.42B as of Q1’26 and 21% increase compared to the $5.95B as of Q2’25.
Share Buybacks and Dividends: $1.3B returned to shareholders through buybacks of $.9B and $.3B in dividends during Q2.
Earnings Call Q&A
The call was split between analysts poking around to see if the $20B fiscal year AI shipment guide was conservative or not, and other analysts finding creative ways to reaffirm the margins would improve in the second half of the year.
Commentary on AI Shipments
Dell asserted in their commentary they are winning deals due to the speed to market their company offers: “Customers are seeing real-time the value in our ability to deploy large-scale clusters quickly and reliability.”
They reiterated they were the first to stand up the GB200 NVL72 and GB300 NVL72s: ‘We were the first in the world to ship both the NVIDIA NVL72 solution last year and the NVL72 system in July.”
When asked right out the gate if the $20B was too low, the CEO stated it was not on their end rather the constraint is the complexity of the systems that are being stood up:
“You've heard us talk about the numbers, but I always sit back and like to reflect on so far through the first half of the year, we've sold $17.7 billion of AI infrastructure. then we shipped $10 billion of that, which would imply we'll ship about $10 billion in the second half equal to the 20. The 5-quarter pipeline continues to grow. — exciting in that pipeline as we saw the sovereign opportunities and the enterprise opportunities grow double digits. But there's complexity here and the complexity lies into these are large-scale deployments. Many have scheduled deliveries and those scheduled deliveries are dependent on things like buildings being ready, power being installed, cooling being installed. — and they are managing a very complex supply chain and a transition as you called out to Blackwall Ultra.”
Margins to Improve by Q4:
The CFO stated the margins would improve by Q4 for increased profitability due to a higher mix of storage and traditional servers.
“From a storage perspective, storage is expected to perform better sequentially in the second half. with more Dell IP as well as normal seasonal acceleration in the fourth quarter. that acceleration in the fourth quarter that storage weighting is what's driving a significant amount of that expected profitability that's implied in our fourth quarter guide. raditional servers are expected to grow in the second half. And of course, we expect our operating expenses to continue to come down as well. So net-net, we expect to be able to deliver more profitability in the second half, and you see that again weighted into the fourth quarter.”
However, for now AI servers are gross dollar accretive and rate dilutive, which we saw this quarter as Dell had a lower gross margin despite adding $500M in gross profits. Dell explained they do foresee their AI server margin expanding as they will have a higher mix of enterprise and sovereign customers compared to cloud service providers (CSPs) where the margins are lower.
“And then if I go back to your question about margins, it's what I tried to articulate earlier with [indiscernible] we expect the onetime costs in our supply chain to reconfigure and to expedite materials not to be in place in the second half. We think there's some opportunity for us to continue to value engineer the scaling of the P&L. And then lastly, the enterprise customers and shipping to enterprise customers and the opportunity to attach unstructured storage, networking and our professional services around that.”
How Low are AI Server Margins, Exactly?
One analyst stated their math implies server margins could be as low as 2% to 2.5% – hence the stock selling off despite the strong AI growth. This is because Dell guided for $4B raise in revenue but only $110M in net profit.
Dell raised their FY2026 revenue from $103B at the midpoint to $107 at the midpoint, with the assumption this from AI servers.
However, Dell only raised GAAP EPS from $7.98 at the midpoint to $7.99 at the midpoint. Adjusted GAAP EPS for the year was raised from $9.40 to $9.55.
This prompted the following exchange, which reveals just how slim the AI server margins are. Notably, the CFO did not deny this calculation.
“Amit Daryanani, Evercore:
I guess I just had a question on the fiscal '26 guide, the way you folks have raised it. You're raising the top line by 4 points or bottom line by about $0.15. It sort of looks like $4 billion more of revenues and about $100 million, $110 million more of net income. And I'm sure there's a lot of moving parts over here, but it almost looks like AI server margins are in the 2%, 2.5% zone for you folks. But maybe just talk about why is the conversion margin so low for the incremental revenues that are coming into the model what are the other puts and takes around it, assuming AI margins are better than that 2.5% matter imply?
CFO:
So if I think about the guide that we have for the second half, certainly, the demand dynamics play a key role in that. So if I think about the traditional server when I think about the AI mix, the biggest impact to the second half and the profitability and outcome is the seasonality within the ISG business and within storage. And so when I think through how we're going to drive more profitability, really do think it's holistic across the board, but it is weighted towards the standard seasonality in the fourth quarter from a storage standpoint. So that's what is embedded within the guide. That's what you can see. That's what we deliver historically, and we — we'll do that again this fiscal year.”
Conclusion:
AI server shipments were a highlight, but AI server margins are a lowlight. That muddies the outcome as typically an investor should be able to celebrate when a portfolio holding reports this kind of AI growth, yet there is very little to celebrate given such thin profitability from the AI segment.
We have a tiny placeholder on this stock of 1% until management can prove there’s an attractive AI business to invest in. We are on the lookout to make sure AI servers have not become a race to the bottom in the competitiveness of winning the Blackwell business.
Should Dell revive its margins, which according to management will be accomplished through the storage attach rate and thier IP portfolio, the stock could become more attractive especially given $20B in AI revenue is nothing to scoff at. Basically, Dell is one to watch but we will not be adding to the position at this time.
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.
Time to power is the most important focal point for AI investors at the moment, considering the high level of demand for compute, the inability of Big Tech to meet said demand in the cloud, and the rapid upgrade cycle between GPU generations.
Once exposed to the volatility and uncertainty of Bitcoin prices and halving, transitioning to offering AI hosting provides more lucrative, predictable, and higher-margin revenue streams for Bitcoin miners. The AI industry is now giving miners a vote of confidence, with up to $50 billion committed in long-term deals this year for AI/HPC data center hosting.
This boils down to two key advantages miners can provide: the ability to retrofit or repurpose existing space for GPUs with lower incremental capex than building AI data centers from scratch, and quicker time to power.
Below, we discuss why time to power is of utmost importance to the AI industry and for AI investors, how Bitcoin miners can bridge the gap and meet near-term power needs, and which BTC mining stocks are engaged in multi-billion dollar AI deals.
The I/O Fund bought shares in Core Scientific early in this trend after first analyzing its AI angle exclusively for our Discovery members. We closed the position shortly after CoreWeave’s acquisition announcement for a 193% return. To access the I/O Fund’s full portfolio, full research library, webinars and more, take advantage of our exclusive Labor Day sale with $110 off Pro and $200 off Advanced.exclusive Labor Day sale with $110 off Pro and $200 off Advanced.
Why Time to Power is Mission-Critical
Power connection is quickly becoming a primary constraint for the AI industry as GPU bottlenecks ease relative to the last two years. Essentially, long lead times for grid interconnection means hyperscalers and specialized AI neoclouds like CoreWeave simply cannot get enough capacity online to meet high demand. This was noted by Microsoft in April and Amazon this quarter, with Amazon CEO Andy Jassy not shy in saying that “the single biggest constraint is power.”
Here's why time to power is now emerging as a critical focal point:
Grid connection timelines in key regions are 36+ months, per data from TD Cowen and Riot Platforms, with major Data Center Alley in North Virginia at 42 months or longer. Even in 2024, Bloomberg reported that utility Dominion said large data centers in Virginia (>100MW) were facing seven year wait times for new connection hookups.
Additionally, the most powerful GPUs from Nvidia (and now AMD) are being upgraded on an annual cadence, so any delays in getting GPU clusters online shortens the time these chips are ‘useful’ before becoming outdated by the next generation. Think of it this way: hyperscalers and GPU providers do not want to spend tens of billions of dollars on AI hardware to then have it sit idle while waiting for power, as this translates directly into lost revenue and lost profits.
Tying into this, if a company like Microsoft can get new data centers connected to the grid and stood up faster than competitors such as Amazon, they can then meet higher demand, win new customers and gain market share. For example, AI startups simply seeking capacity to train models likely have lower cloud switching costs, meaning that these companies could easily and quickly shift workloads to whichever cloud provider can offer them the capacity they need the quickest.
Bitcoin Miners Provide a Path to Meet More Immediate Power Needs
While some companies like Meta are building 1GW+ clusters from scratch, for others in the neocloud arena that do not have access to tens of billions in cash, Bitcoin miners are recently attracting increased levels of interest. Miners help address immediate power needs with multiple gigawatts of grid-connected facilities that can quickly be converted from mining to AI GPUs. As TeraWulf’s executives put it, the “ability to scale quickly provides a meaningful advantage in today’s race to secure power and compute capacity.”
Consider some of those challenges that exist on the power side when building data centers from the ground up – short power supply in key markets like Virginia, grid connection requests extending four to seven years, or permitting delaying full power delivery to new infrastructure. Miners, on the other hand, streamline this process by offering access to readily available grid-connected power, cooling infrastructure, and low latency to major hubs.
Core Scientific, Applied Digital and Galaxy are expecting to bring their first AI-focused facilities online by Q4 2025 through 1H 2026, or four to ten months from now, substantially quicker than new construction. Transitioning existing facilities also offers rapid time to operation; for example, IREN stated that it seamlessly transitioned from mining ASICs to AI GPUs at its 50 MW Prince George facility in six to eight weeks, though this is likely at a very small scale considering the size of its GPU fleet.
For a company like CoreWeave or Fluidstack, signing long-term contracts for capacity with BTC miners and fronting the cash for capex allows them to build hyperscale clusters for a fraction of the cost and time. As CoreWeave’s CEO Michael Intrator put it, “Right now the key bottleneck really is the powered shell. When you think about that, that’s really the building, the cooling, the distribution of electricity.”
Instead of having to worry about the prolonged process of site selection, permitting, planning, and more before final power delivery, these neoclouds instead have near immediate access to the powered shell. While retrofitting for liquid cooling, networking and connectivity may still be necessary and pose some challenges, miners offer a fast time to operation and relatively cheaper capex costs for a hyperscale-size data center outfitted with tens to hundreds of thousands of GPUs.
Considering that most of the miners that have struck deals are rather cash-strapped, these GPU cloud firms behind the deals are putting up all (or the majority) of the capex required to retrofit existing facilities for high-end AI GPUs. Miners are highlighting the capex costs per MW for these facilities, and these costs, while varying quite wildly between miners, help emphasize the industry’s rather attractive positioning when factoring in quick time to power.
Core Scientific had offered one of the cheapest project costs, costing CoreWeave approximately $1.5 million per MW, though Core Scientific’s executives had stated in Q3 2024 that “they were able to significantly buy down their rates.”
IREN is touting the next lowest capex costs at ~$6-7 million per MW of IT load, for its Horizon 1 facility in Texas, which it says is catered to liquid-cooled NVIDIA Blackwell GPUs and provides ~6ms latency to Dallas. IREN says it expects the capex to be funded primarily by colocation customer prepayments and debt financing.
TeraWulf expects its capex per MW of critical IT load to be in the range of $8-10 million for its recent deal with Fluidstack, while offering <8ms latency to New York City and Boston and <2ms to Toronto.
Applied Digital and Galaxy are a bit on the higher end – both are anticipating capex of $11-$13 million per MW, with Galaxy recently securing $1.4 billion in project financing to aid development of its first 133MW phase with CoreWeave. However, Galaxy expects capex to move higher for its second 260 MW phase.
Compared to new construction costs per MW, miners can offer some capex savings, with savings becoming more attractive the lower capex costs go, such as in the case of IREN and TeraWulf. Cushman & Wakefield estimates construction costs per MW in the range of $9-15 million across key markets, averaging $11-13 million. This aligns with estimates from CBRE for $10-14 million per MW, though costs can reach $16-20 million per MW in certain cases (or higher).
While Applied Digital and Galaxy may not offer much of a discount relative to new construction, IREN and TeraWulf can offer more pronounced capex savings. Based on a $12 million per MW construction cost estimate, TeraWulf could offer up to 33% savings per MW and IREN up to 50% savings, translating into hundreds of millions to billions in savings at a 400-800 MW scale.
Time to Power Comparison
Theoretically speaking, miners with the quickest time to power should enjoy a two-fold advantage: quicker time to revenue recognition, and potentially better deal economics arising from the ability to meet customers’ demand faster.
For example, Core Scientific’s ability to bring 250MW of power online for CoreWeave’s AI compute needs supports why it was the first to be acquired, as no other miner could match that scale within the same timeframe.
There are nuances in delivery timelines and expectations that make it difficult to offer a direct quarter-by-quarter comparison, such as Galaxy only offering vague delivery comments for 1H or throughout the year. However, roughly speaking (and excluding Core Scientific), by the end of 2025, Applied Digital should lead with 100MW, yet this shifts by year-end 2026 with TeraWulf aiming to deliver more power.
By Q4 2025:
Applied Digital expects its first 100MW of power online.
TeraWulf expects 60MW of power online.
IREN could have 50-130MW of AI-dedicated power online, depending on how it retrofits its Prince George and Mackenzie facilities.
Galaxy does not expect to have capacity online.
By Q4 2026:
TeraWulf expects to have 420MW of capacity online as it progresses with quick expansion.
Applied Digital expects to have 250MW of capacity online.
Galaxy expects to have 133MW of capacity online.
IREN is aiming to energize its 1.4GW Sweetwater facility in early 2026, though there is no indication yet that it will be able to stand up fully operational infrastructure by year-end.
Long-Term Power Needs Far Outstrip Miner Capacity
Looking at the bigger picture, the contracted power that miners can provide will only meet a small portion of the projected demand growth over the next three to ten years, and are unlikely to be the sole solution to this structural power shortfall. As we had covered in our free newsletter, Nuclear Power Emerging as a Clean AI Data Center Energy Source, data center power demand is expected to grow at an accelerated 16% CAGR from 2023 to 2028 and beyond.
For example, Boston Consulting Group forecasts 45 GW of growth in global data center power demand in just three years, from 82 GW in 2025 to 127 GW by 2028.
In the US, data center demand was forecast to rise as much as 5x over the next decade. Deloitte estimated US data center power demand to triple from 2025 to 2030, from 41 GW to 120 GW, before rising further to 176 GW by 2035. In terms of power consumption, the DOE recently forecast data centers will consume 6.7% to 12% of total US power production by 2028, up from 4.4% in 2023.
Bitcoin miners can only meet a fraction of this growth, likely around several gigawatts in total. Yet their innate ability to deliver this power over the next 12 to 24 months, supporting up to hundreds of thousands of high-end GPUs in larger-scale facilities, is why miners are prime targets to meet hyperscalers and neoclouds’ immediate power needs.
Below, we discuss miners that have secured large-scale deals and/or progressing with building out GPU fleets for self-hosted AI cloud services. As it stands, CoreWeave has been the primary undertaker of major AI deals with miners, committing to over $35 billion including its acquisition of Core Scientific; though Fluidstack is making a splash with a deal worth up to $16 billion with TeraWulf.
Core Scientific Acquired by CoreWeave Following $10B+ Deal
CoreWeave made a statement to the industry with its acquisition of CoreScientific for $9 billion, a Bitcoin miner and key compute partner. The acquisition communicates that immediate power at scale is paramount. This stems from Core Scientific’s key advantage, as the only miner that will be able to bring 250MW online this calendar year, capable of hosting >100,000 GB200 chips.
Prior to the acquisition, CoreWeave had contracted 590MW of total capacity from Core Scientific as of February, worth $10.2 billion in cumulative revenue to the miner over the 12-year lease terms. Core Scientific stated in May that it was on track to deliver 250MW of billable capacity to CoreWeave by the end of 2025, with expectations for the full 590MW to be delivered by 2027.
CoreWeave says that with the acquisition, it will now own 1.3 GW of gross power across Core Scientific's national footprint (including BTC mining facilities), with an incremental 1 GW+ of potential gross power available for future expansion. The acquisition also adds $500 million of estimated fully ramped, annual run rate cost savings by eliminating CoreWeave’s $10 billion lease obligations to Core Scientific.
This combination of hundreds of MW of deliverable power by year end and substantial long-term cost savings from owning infrastructure outright versus leasing underscore why CoreWeave was quick to progress with the acquisition.
TeraWulf Signs up to $16B Deal with Fluidstack, Backed by Google
TeraWulf is no stranger to AI/HPC hosting, having signed a $1 billion, 72.5 MW deal with Core42 in December 2024, with dedicated GPU compute infrastructure coming online throughout 2025. However, its recent deal with Fluidstack represents one of the largest deals to date between an AI cloud firm and a Bitcoin miner.
TeraWulf had originally announced a $3.7 billion, 10-year AI hosting deal with AI cloud startup Fluidstack for 200+ MW capacity at its CB-3 and CB-4 buildings at its Lake Mariner data center in upstate New York.
Only a few days later, TeraWulf announced that Fluidstack was expanding its deal to include ~160MW at CB-5, bringing its total contracted IT load to 360MW and contracted revenue to $6.7 billion. Including potential lease extensions, the new deal could be worth up to $16 billion.
Perhaps more importantly, Google is backstopping $3.2 billion of Fluidstack’s lease obligations to support project financing, and has taken a 14% stake in the miner (up from a $1.8 billion backstop and 8% stake with the original deal).
TeraWulf is aiming for quick delivery of power to Fluidstack, with the first 40MW phase expected to come online in the first half of 2026 and all 200+MW delivered by year-end; additionally, operations are expected to commence for the 160MW tranche at CB-5 in 2H 2026. Multiple different OEM GPU equipment is expected to be deployed, likely GPUs from Nvidia and AMD, able to serve a variety of AI training and inference workloads.
Overall, Fluidstack’s deal represents just over 70% of the current available power at Lake Mariner of 500MW, though TeraWulf does have 250MW additional power pending regulatory approval for expansion. This would bring the facility’s total power potentially up to 750MW with its targeted upgrades, or allowing a second deal of similar size to be signed in the future.
TeraWulf also recently signed an 80-year ground lease with purchase option at Cayuga, securing its exclusive rights to develop 400MW of data center infrastructure “on a fully equipped site, with high-capacity transmission, industrial water intake and redundant fiber.” The company expects to bring the first 130MW online in 2027, while boosting its long-term power capacity to over 1GW.
Applied Digital a Secondary Benefactor of CoreWeave with $11B Deal
If CoreWeave’s acquisition of Core Scientific wasn’t enough, it also committed to leasing 400MW of capacity from Nvidia partner Applied Digital at its Ellendale, North Dakota facility, worth ~$11 billion over the 15-year terms. Ellendale (now referred to as Polaris Forge 1) has the potential to scale up to over 1GW capacity over the long run, though the 400MW represents the maximum current capacity.
Applied Digital expects the first 100 MW data center to be ready for service in Q4 2025, while the second 150 MW data center is expected to be ready in mid-2026. The third 150 MW facility is expected to be ready in 2027. Applied shared some details on the campus, saying that the campus “will feature high-density racks and direct-to-chip closed-loop liquid cooling and air cooling combo” as it entered a $150 million, 36-month convertible preferred equity financing to advance construction.
On August 18, Applied announced that it is expecting to break ground on its $3 billion, 280MW Polaris Forge 2 data center in Harwood, North Dakota. Applied is aiming to have initial capacity online in 2026 and the facility reaching full capacity by early 2027, a quick 16-20 month time to power. The company also claimed in 2024 to offer approximately 30% lower costs than AWS, Azure and GCP for cloud compute on Nvidia’s GPUs with short, 8-10 week lead times, a key advantage in the race to deliver compute.
Galaxy: Another CoreWeave Partner with ~$15B Deal
CoreWeave has struck another deal with digital asset platform and now data center infrastructure provider Galaxy Digital, committing to the entire 800 MW of approved capacity (~526MW critical IT load) at Galaxy’s Helios data center. With 2.7GW of power under load study, Helios has the potential to expand up to 3.5GW, which would make it one of the largest single data center facilities in the US and the world.
Galaxy stated that they anticipate average annual revenue of more than $1 billion over the 15-year term, based on committed contractual terms, internal capex estimates, and full capacity utilization. This estimate would place the deal value above $15B.
Galaxy is expected to deliver its first 133 MW phase of power to CoreWeave in the first half of 2026, followed by the next 260 MW phase throughout 2027 and a subsequent 133 MW phase throughout 2028.
To fund this first phase, Galaxy has secured $1.4 billion in project financing debt, providing the $350 million equity requirement for the funding. Management also stated in Q2 that its $480 million in cash proceeds from its May equity raise would go towards capex related to the Helios DC buildout. Overall, the combined phases are no small task, likely requiring close to $10 billion over the next few years, especially considering management stating Phase 2’s will likely be slightly higher than Phase 1 due to the size of the committed load.
Galaxy made it clear that for Phase 2, discussions for project financing are still “pretty preliminary,” though management expects that as they produce results with Phase 1 and generate returns/revenue, they will “earn the right to achieve larger financings at lower cost.” Simply getting the second phase financed is the company’s primary goal, considering the 16-20 month timeline to delivery.
IREN Building Out its Nvidia GPU Fleet, Targeting Deals
In March, IREN hit the pause button on its bitcoin mining expansion as it pivoted to focus more on AI and HPC, given the revenue potential stemming from its >2.9 GW of grid-connected power.
Unlike peers, IREN is primarily focusing on short-duration contracts, from on-demand use to three-year term lengths, while expanding its GPU fleet to drive growth in its self-hosted AI cloud business. Revenue for its AI Cloud Services surged 94% QoQ to $7 million in fiscal Q4, after rising 33% QoQ to $3.6 million in its fiscal Q3. The company laid out an aggressive $200-250 million annualized AI Cloud Services revenue target by December 2025, up 8-10x versus its current annualized run rate.
IREN is working to expand its GPU fleet, which is still extremely small considering its power pipeline as GPUs are not cheap at scale. IREN purchased 2,400 Blackwell GPUs in early July for ~$130 million including fit-out costs, comprised of 1,300 B200s and 1,100 B300s.
In late August, IREN doubled its fleet to ~8,500 GPUs with another purchase of 4,200 B200s for ~$193 million, while securing $102 million in funding for the July purchase. The company has already contracted out its first batch of 256 B200s to an undisclosed customer.
In its fiscal Q4 report, IREN also stated that it has secured $200 million in GPU financing, aiming to increase its GPU fleet to ~10,900, or an additional ~2,400 GPUs.
Given that a 50MW data center could be outfitted with ~25K GPUs, IREN is far from outfitting its own facilities with its current fleet, though it is somewhat capital constrained given its $564 million cash on hand.
The company says that it is “observing demand for multi-thousand air-cooled Blackwell GPUs,” and has ~47MW of capacity still available at its Prince George, British Columbia facility, capable of supporting ~20,000 Blackwell GPUs.
For its other AI data centers, IREN is targeting calendar Q4 2025 delivery at its Horizon 1 facility in Texas with up to 50 MW of IT load. IREN says that it has several customers undergoing due diligence and contractual negotiations with interest expanding beyond 50MW, though it has not announced any signed deals. The company is also eyeing a complete transformation for Horizon to reach the full 750MW power capacity, noting in fiscal Q4 that procurement is underway for a second 50MW.
IREN is looking to create an interconnected 2GW data center hub at its Sweetwater 1 and 2 facilities, with the 1.4GW Sweetwater 1 facility expected to be energized by April 2026 and the 0.6GW Sweetwater 2 facility expected to be energized in late 2027. This is a slight acceleration for Sweetwater 2 from early 2028. Together, Sweetwater could support 700K liquid-cooled Blackwell GPUs, in close proximity to Stargate’s Abilene data center.
Other Miners Considering AI Pivots
Riot Platforms and Hut 8 are two other miners that are pivoting towards AI hosting, though the two have yet to secure long-term contracts.
Riot is aiming to transition from Bitcoin mining to AI hosting when “economic and feasible,” and said last quarter that it is “actively progressing toward securing a lease with a high-quality tenant” at its Corsicana facility. As of January, Riot was evaluating using the remaining 600MW of power capacity at the facility for AI/HPC, with 400MW currently geared towards mining. Needham analysts are encouraged by Riot’s possible pivot, saying they “believe Riot’s Corsicana site is one of the most attractive HPC sites in our universe” with amplefiber connectivity for low latency AI inference. Needham believes Riot could be in advanced discussions with potential tenants by 2H 25 and sign a lease as early as Q1 2026.
Hut 8 has ~3.4MW of HPC capacity operational and ~670MW of mining capacity, and similar to IREN is leasing Nvidia GPUs in the cloud. Hut reported a $2.3 million increase in revenue from leasing ~1,000 H100 GPUs to an unnamed AI developer, which launched in September 2024. Hut 8 also just broke ground on a ~300MW data center in Louisiana, and is expected to invest ~$2.5 billion, while undisclosed future tenants are expected to contribute ~$10 billion in compute equipment. However, it is reported that Hut will likely need external financing or a partner for the project.
Earlier in 2025, Bitfarms announced it was mulling a shift to AI, and in early August, the miner announced a partnership with T5 Data Centers to advance the development of its Panther Creek facility in Pennsylvania. The two are expected to engage in pre-construction design planning and development approval processes, with Bitfarms also submitting its master site plan to Macquarie for future development.
Cipher is also considering a pivot, saying that it created a new strategic plan at its 150MW Black Pearl Phase II to bridge both AI compute and hydro-bitcoin mining, though in the long run it expects the site to be fully leased by AI/HPC tenants. Cipher added that it is seeing continuing HPC interest at its Barber Lake site.
AI/DC Deals are High-Margin, Highly Visible Revenue for BTC Miners
These AI hosting deals are very attractive for miners as they provide highly visible, high-margin revenue streams, a major operational shift from the prior business model with growth and earnings tied to Bitcoin’s price volatility, network difficulty and halving.
Applied Digital’s deal with CoreWeave has pricing set upfront with an annual escalator, with average annual revenue of ~$733 million based on the $11 billion, 15-year terms. Revenue is likely to start small and begin to ramp rapidly once full capacity is online. Applied is also targeting 88%, +/- 3% net operating income margins on the AI hosting revenue, or ~$645 million average annual net operating income for the duration of the deal.
TeraWulf’s expanded deal with Fluidstack will generate average annual revenue of $670 million at its current scope, though the full extension to $16 billion could provide significant revenue upside. TeraWulf is eyeing an 85% net operating income margin for the Fluidstack deal, or ~$570 million on average annually. This is at the low end of Applied Digital’s range, likely accounting for higher operating costs due to location (upstate New York vs North Dakota). Combined with the Core42 deal, TeraWulf has visibility into ~$770 million in average annual revenue, with the first revenue from Core42 now being recognized.
Galaxy is targeting 90% adjusted EBITDA margins on its AI hosting deal with CoreWeave, implying adjusted EBITDA of ~$900M+ on its average annual revenue estimate of more than $1 billion. Given energization times spanning into 2028 and a longer 15-year term structure, Galaxy’s initial revenue ramp may be more prolonged than peers.
IREN does not have a firm hosting deal, though it has touted a 97-98% hardware profit margin for its AI Cloud Services, or revenue minus electricity costs. This is more than 20 points above its Bitcoin hardware profit margin, highlighting why AI is more attractive than mining at scale. However, margin tailwinds are likely minimal in the near-term given AI Cloud Services contributes less than 3% of revenue.
Accelerated Revenue CAGR
While there are nuances in deal sizes and lengths, IREN, Applied Digital and TeraWulf are seeing accelerated forward revenue growth CAGRs as AI capacity soon comes online. These miners’ growth rates are much stronger than others such as Riot and MARA who have not jumped headfirst into AI hosting.
Applied Digital and TeraWulf are expected to see revenue increase at an 88-90% 2-year CAGR from 2025 to 2027, a significant acceleration from their respective historical 61% and 73% CAGRs from FY23 to FY25.
IREN is expected to see revenue grow at a 42% CAGR from FY25 to FY27, though this is a bit skewed as the company has benefitted from rapid mining hashrate expansion and rising Bitcoin prices over the past two quarters.
Compare this to Riot and MARA, with both expected to see FY25 to FY27 revenue growth at a 17-18% CAGR, decelerating sharply from the 55-60% level over the past two years. While revenue is coming off a higher base than say APLD or WULF, the difference in forward growth rates is notable for AI-engaged miners and these two who have not yet pivoted in full force.
Thin Balance Sheets Present Capital Raise Risk
Miners are rather cash-strapped, and while neoclouds and partners are fronting the cash for capex, capital raises and dilution are still a risk, considering IREN and TeraWulf both recently launched larger-scale convertible note offerings.
Here’s a quick snapshot into the health of IREN, TeraWulf, Applied Digital, and Galaxy’s cash versus debt, with the chart below showing how thin and lumpy cash balances are:
IREN reported cash and equivalents of $196 million in Q3, down from $455 million in the prior quarter. As of Q4, IREN reported $564 million in cash and equivalents, after closing an upsized $550 million convertible note offering in June, while debt is now at $963 million. This presents possible dilution risk in the high-teens, based on IREN’s $5 billion valuation.
Applied Digital reported cash and equivalents of $114 million, down from $254 million in the prior quarter. Applied has a deal with Macquarie for up to $5 billion in financing, including a $900 million initial investment at Ellendale and $4.1 billion on retainer to for future data center expansion. For any future builds, Macquarie would invest $2.25 million per MW and Applied would invest $0.75 million per MW.
TeraWulf reported cash and equivalents of $90 million, down from $220 million in the prior quarter. However, shortly after the Fluidstack deal, the company announced the full exercise of its $1 billion convertible note offering, or ~27% of its current market cap. Based on prior capital allocation projections, this would likely leave TeraWulf with ~$600-700 million in unallocated cash, for project cost overruns or other expansion needs.
Galaxy reported cash and stablecoins of $1.18 billion in Q2, including $691 million in cash and $489 million in stablecoins, approximately flat from the prior quarter. Notes and loans payable were $1.07 billion.
Galaxy President and CIO Christopher Ferraro offered some very important perspective on funding and capacity growth, and why capital is likely to be the limiting factor for miners’ current buildouts:
“There’s also a practical component, which is, these are very large-scale, long-term development projects that take a lot of capital. And so our ability to grow into the opportunity is wholly dependent on 2 things: one, us executing excellently; but then also two, growing and getting bigger as a company so that we can actually support the growth, meaning like it would be totally imprudent for us to now take on in parallel, for example, like another $10 billion build, because that requires a capital base and the attention and resources that we're just not built out for today.”
This matches our statements in our Core Scientific analysis from May, Core Scientific Q1: Expects 250MW of Billable Capacity to CoreWeave by Year-End, where we explained that if CoreWeave did not front the capex for the data centers, the “business model would not work as CORZ would struggle to raise the level of capital required to acquire more sites and modify the existing infrastructure.”
These comments that the current builds underway for CoreWeave and Fluidstack are likely to be the main focus of the miners over the next few years, especially considering the thin balance sheets, convertible note raises, and difficulty from CoreWeave to keep funding multiple different projects.
Customer Concentration Another Risk to Consider
One other risk to consider is customer concentration, given the fact that CoreWeave is the sole backer for a majority of the miners discussed here. TeraWulf has the benefit of Google backstopping Fluidstack’s obligations, offering early termination protection for the first six years.
Considering CoreWeave has made the move to acquire Core Scientific and has signed deals with Applied Digital and Galaxy, it may be more limited in its ability to fund future projects. It also means that miners signing away all or a majority of their power to CoreWeave find it hard to diversify revenue streams away, with Galaxy noting that the deal with CoreWeave “is going to take up the vast majority of our attention over the next few years” and prevent other hyperscaler engagements.
Also, if CoreWeave should pull away from a deal down the road, it could create a significant blow to revenue and earnings for miners it is currently engaged with. This is due to two factors: high-margin, high concentration of revenue CoreWeave’s deal will drive, and that there is little room for diversification as new capacity is expensive for miners to handle without major financing.
Technical Analysis
The risk associated with Bitcoin miners is present in the potential setups outlined below. As you will see, some have the potential for wild swings in either direction, which follows very messy and overlapping uptrend patterns. For this reason, we approach these charts only from the mindset of risk management. If we do take a position in any of these names, we do not view them as buy-and-hold vehicles, and all will come with relatively tight risk controls.
Galaxy Digital (GLXY)
Green – GLXY appears to be tracing a very large diagonal pattern. If we zoom into the current drop in GLXY, it appears to be a 4th wave, and needs one more swing higher to complete the larger pattern. The target for this swing is $34 – $43. As long as we hold $19, this setup remains valid.
Blue – Instead of getting one more swing higher, we should see two more. Once we reach $35 – $43, we’ll see another 3-wave drop and a final swing to $56 – $67. Once again, if we break below $19, then both of these scenarios are no longer valid, and a larger top will likely be in.
TeraWulf (WULF)
Blue- We have completed wave 1 in a very large diagonal, as well as wave 2. We are just now completing wave A of 3. This should be a double top that turn lower in 3 waves toward $5.25 – $3.65. The drop needs to be a 3-wave drop to confirm this count. Once this ends, we should see a 5-wave pattern turn higher. This would point toward $69 – $105 in a pretty direct path.
Green – If we instead keep pushing higher over $12.50, with volume and momentum expanding with price, then we could be in a more direct path higher. Instead of a large diagonal pattern, it would be a standard 5 wave pattern. Wave 1 of 3 would push toward $35 – $69.
Please note, while the pattern below allows for these counts, we still only have 3 waves up off the 2022 low. So, until we see either the green count get confirmed, or a 3-wave correction, risk remains high. Any drop below $4.20 will be the first warning. If we see a 5-wave drop break below this level, the risk will increase that something more bearish might be in play.
Applied Digital (APLD)
Green – APLD is tracing a very large 3 wave pattern, which best fits as a diagonal pattern. If we can see a breakout on heavy volume and expanding momentum over $17.25, it will confirm this count, which should see a continuation towards $23 – $29.
Blue – We fail to breakout over $17.25 and instead drop back into the $14.50 – $12.50 range. We’ll then break through the $11.35 region and head toward $9 in a larger 2nd wave.
Riot Digital (RIOT)
Blue – We are completing the B wave in a larger 2nd wave. We will fail under $14.10 – $15.35 and see a sharp, 5-wave drop back to $9 – $8. This will hold, as we set up for a larger 3rd wave breakout to new highs.
Green – We completed the 2nd wave, and it was shallow. We’ll see a strong breakout over $14.10 – $15.35 on expanding volume and momentum. the 3rd wave target is $27 – $31.
Both counts must hold $7.56 on any weakness, or they will get invalidated.
Iren Limited (IREN)
Green – Note how the most recent push higher is happening on less volume and momentum. This is the 5th wave, which can continue as high as $35. Once complete, this will end a large 3rd wave within a diagonal pattern. The target for the 4th wave is around $8. We should then see a large 5th wave to new highs.
Red – We only have 3 waves up off the 2022 low. This is not ideal, and until we see a 3 wave retrace, it is a risk that should not be ignored. If this is all we get, then we should see a large 5 wave drop develop that takes us through $8 and then $2.80. We would then be on our way to new all-time lows.
While both counts have IREN in a 5th wave, if we can see volume and momentum expand with price over $35, then something more bullish may be going on. If we see this then we will adjust accordingly.
Conclusion
Bitcoin miners are not able to solve the long-term power deficit that the industry is coming head-to-head with, but in the 12 to 36 month window, miners are uniquely positioned to meet hyperscale and neocloud power needs. AI hosting deals provide highly-visible, high-margin revenue over the next decade and beyond, a major operational shift for miners once reliant on volatile Bitcoin prices for growth.
The catch is that miners cannot finance these AI data center builds themselves, as they have thin balance sheets and larger debt loads. Instead, they are reliant on their backers such as CoreWeave, and in TeraWulf’s case, Fluidstack and Google, to front the cash for the projects and provide the necessary compute.
While risk does stem from the fact that CoreWeave is the sole backer for a majority of these AI hosting deals, limiting opportunities for revenue diversification and minimizing concentration risk, AI presents much stronger, higher-margin forward growth opportunities.
Damien Robbins, Equity Analyst at I/O Fund contributed to this analysis.
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.
In late 2022, Bitcoin fell into the $16,000 range amid the fallout from the FTX scandal. At the time, credible institutional buy calls were nowhere to be found, as Wall Street largely stayed on the sidelines.
Fast forward to today—Bitcoin now trades more than 600% above its 2022 lows, and a flood of institutional optimism has now entered the news cycle. Analysts and money managers are now calling for another doubling in price before year-end.
Market sentiment has always been a potent force. History shows that the most enthusiastic bullish narratives tend to emerge when prices are stretched, while more cautious takes often surface near market bottoms ,and this cycle is no different.
Throughout Bitcoin’s history, the I/O Fund has continued to distinguish itself with timely, accurate Bitcoin analysis dating back to 2019. While others chased the 2021 frenzy with $200,000 to $500,000 price targets, we remained disciplined—scaling back crypto exposure by half to secure profits. Later, when Bitcoin revisited the $16,000 region, we issued a Strong Buy Alert to our free subscribers, which was subsequently picked up by Tier 1 media.
Our track record is not the product of hype but of a systematic framework—one built on technical analysis, on-chain metrics, and a close watch on global liquidity conditions. Today, this very process is flashing warning signs, which has us, once again, going against the popular narratives, as we have begun the process of taking gains in Bitcoin and reducing risk.
Because of the warnings that our system is picking up on, the I/O Fund partnered with WealthUmbrellato release a free one-hour webinar outlining these risks as well as what we want to see to confirm a potential path higher from here. The presentation details the forces that move Bitcoin, along with the risk management techniques that enabled us to sidestep the 2021 top and re-enter near the 2022 low.
Below are a few highlights from the presentation:
In the below clip, Knox Ridley highlights the unique nature of Bitcoin’s price movements and how the I/O Fund manages risk in a market where narrative-based investing does not work. Since Bitcoin lacks traditional fundamentals, we rely on an original process designed specifically for crypto — one that has allowed us to repeatedly identify both tops and bottoms.
Though classic fundamental analysis does not apply to Bitcoin, there is a unique set of data points that can be analyzed to help determine the health of Bitcoins’ trend. This type of analysis is called on-chain analysis, which is explained in the below Clip, by Vincent Duchaine of WealthUmbrella.
Sign up below to Access the full video, which will offer:
Interactive charting of Bitcoin.
The price levels that must hold to confirm another leg higher in Bitcoin.
What our game plan is and how far Bitcoin will drop if this level breaks.
What on-chain analysis is telling us about the health of Bitcoin.
The probable price target that WealthUmbrella is projecting before seeing a cyclical top.
If you are a crypto investor who would like know our plans for participating in any remaining upside Bitcoin has to offer, and how we plan to further minimize the downside, we encourage you to attend one of our weekly webinars. Every Thursday at 4:30 EST, portfolio manager, Knox Ridley, goes through various broad market charts, as well as discusses our game plan on various stocks and crypto currencies that we currently own or want to own. Learn more here
Please note: The I/O Fund conducts research and draws conclusions for the company’s portfolio. We then share that information with our readers and offer real-time trade notifications. This is not a guarantee of a stock’s performance and it is not financial advice. Please consult your personal financial advisor before buying any stock in the companies mentioned in this analysis. Beth Kindig and the I/O Fund own BTC at the time of writing and may own stocks pictured in the charts.
Nvidia is the global AI compute leader and there is not a distant second. Therefore, reporting (1%) for Q2’s compute segment should have tanked the stock. Instead, we are seeing a mild reaction because Q3 was quite strong and spells good things to come for Nvidia in the second half of the year.
Given we are on the cusp of Blackwell starting to drive the stock’s narrative (c’mon already), it’s a good time to pause and talk about where Nvidia could go from here. I revisited the medium-term and longer-term forecast on a few pre-earnings discussions and was pleasantly surprised that Jensen Huang did the same on the earnings call.
Also important for I/O Fund members who hold AI networking positions, the report was much stronger than expected with networking up a whopping 46% QoQ and 78% YoY to $7.25 billion. This foreshadows what’s to come in the AI networking space; and should be a boon for stocks like Astera Labs and Credo which supply Nvidia.
Despite Q3 putting Nvidia right on track, I would not be surprised if the market offers soft price action on the stock in the near-term. The China narrative is confusing (and beaten to death), Q2 should mark a bottom and semiconductors don’t do well at their cyclical bottoms, plus the AI market is running on fumes until Nvidia can carry the market (Q3-Q4).
First, we will run through the financials, and then in the second section I’ll break down what I think you can expect on this stock for the remaining part of this year, into calendar year 2026 and by the end of the decade.
Slight Revenue Beat in Q2, Solid Q3 Guide
Nvidia reported $46.74 billion in revenue in Q2, slightly ahead of estimates for $46.13 billion. This corresponded to growth of 55.6% YoY, decelerating more than 13 points from 69.2% in Q1; on a QoQ basis, revenue increased just 6.1%, slowing from 12% in Q1.
This is expected to be a temporary lull in sequential growth, as Nvidia guided for $54 billion in revenue, +/- 2%, for Q3, corresponding to 53.8% YoY growth and a rebound to 15.5% QoQ growth. This was ahead of estimates for $52.7 billion, even with Nvidia noting that this guidance assumed no H20 sales in the upcoming quarter.
For FY26, Nvidia is currently expected to report 55.4% revenue growth to $202.8 billion, though it is likely that this figure is revised higher, towards $205 billion, accounting for the slight beat in Q2 and Q3’s guide.
Key Segments: First Sequential Decline in Compute Revenue, No China Sales
Nvidia’s data center revenue increased 56% YoY and 5% QoQ to $41.1 billion, a marginal miss versus estimates for ~$41.2 billion in the quarter. This is also the smallest sequential increase since Hopper’s breakout quarter at just ~$2 billion.
Notably, Q2’s data center result would mark the segment’s first miss since Q4 FY23 as Q1 was essentially in-line.
While Compute revenue was up 50% YoY, the segment reported an unusual (1%) QoQ decline to $33.8 billion, a sharp shift from double-digit growth in late FY25 and 5% in Q1. Nvidia said the sequential decline was driven by a $4 billion QoQ reduction in H20 sales, as it reported no H20 sales to China this quarter and $0.65 billion in H20 sales to an unrestricted customer outside of China.
On a more positive note, Nvidia said that it is continuing to ramp Blackwell and now Blackwell Ultra GPUs, with growth of 17% QoQ. Based on comments from Q1 placing Blackwell revenue in the $23.5-24B range, this would place Blackwell revenue approaching $28 billion.
Networking growth was also robust in Q2, helping offset the softness in Compute as revenue rose 46% QoQ and 78% YoY to $7.25B. This marked a 22 point acceleration from 56% YoY growth in Q1. Nvidia said the strong performance in Networking was driven by “growth of NVLink compute fabric for GB200 and GB300 systems, the ramp of XDR InfiniBand products, and adoption of Ethernet for AI solutions.”
Gaming revenue was $4.29 billion, rising 14% QoQ and accelerating 7 points to 49% YoY. Nvidia said this was driven by sales of Blackwell-based GPUs.
Automotive revenue rose 69% YoY and 3% QoQ to $586 million.
Pro Viz revenue rose 32% YoY and 18% QoQ to $601 million.
OEM and other revenue rose 97% YoY and 56% QoQ to $173 million.
Margins Outperform in Q2 with Strong Expansion
Despite the softness in data center compute, Nvidia outperformed on margins, beating its guidance across the board and delivering a notable uplift in operating margin. Nvidia guided for more sequential improvement in operating margin in Q3, to the highest level since Q1 FY25.
GAAP gross margin was 72.4% in Q2, slightly ahead of guidance for 71.8%. Adjusted gross margin was 72.7%, or 72.3% excluding $180M in H20 inventory releases, ahead of guidance for 72%.
For Q3, Nvidia guided for GAAP gross margin to improve nearly 1 point to 73.3%, +/- 0.5%, and adjusted gross margin to improve to 73.5%, +/- 0.5%.
GAAP operating margin was 60.8% in Q2, improving nearly 12 points QoQ and coming in 1.7 points ahead of guidance for 59.1%. Adjusted operating margin was 64.5%, also up nearly 12 points QoQ and 1.4 points ahead of guidance for 63.1%.
For Q3, Nvidia guided for GAAP operating margin to expand 1.6 point QoQ to 62.4%, signaling continuing operating leverage tailwinds in the quarter. Adjusted operating margin was guided at 65.7%, up 1.2 points QoQ.
GAAP net margin was 56.6%, up 14 points QoQ and more than 6 points ahead of guidance for 50.2%. Adjusted net margin was 55.2%, up 10 points QoQ.
For Q3, Nvidia’s guidance implies a GAAP net margin of 52.9%, moderating 3.7 points QoQ.
Slight Adjusted EPS Beat in Q2
Nvidia reported just a 3.9% adjusted EPS beat in Q2, reporting $1.05 in earnings versus estimates for $1.01. This corresponds to growth of 54.4% YoY, rebounding substantially from Q1’s H20-affected 32.8% growth. Adjusted ESP growth is expected to remain strong through the rest of the fiscal year, at 47.2% and 53.2% in Q3 and Q4.
For FY26, analysts are currently expecting Nvidia to earn $4.36 in adjusted EPS, up 45.8% YoY. This has improved $0.20 since May’s estimate of $4.16, or 39.2% YoY growth. While still early, Nvidia’s earnings growth is expected to remain quite strong in FY27 at 38% YoY to $6.02.
Cash Flow Margins Drop to Lowest Level since Q4 FY23
Nvidia’s cash flow margins dropped to the lowest level since Q4 FY23, as sharp QoQ growth in accounts receivable and inventories weighed on both operating and free cash flow.
Operating cash flow was $15.37 billion, down from $27.41 billion in Q1. OCF margin was 32.8%, down nearly 30 points QoQ and more than 15 points lower YoY.
Free cash flow was $13.45 billion, down from $26.14 billion in Q1. FCF margin was 28.8%, again down more than 30 points QoQ and more than 16 points lower YoY.
Accounts receivable surged $5.7 billion QoQ, or nearly 26%, to $27.8 billion. Nvidia said this was driven by timing of cash collections and Blackwell Ultra ramping late in the quarter.
Inventories rose more than $3.6 billion QoQ, or 32%, to $14.96 billion, to support Blackwell Ultra’s ramp.
Cash and equivalents totaled $56.8 billion, while debt remained steady at $8.47 billion.
$200 Billion Run Rate (and why it matters)
The Q3 guide was key as if we assume a similar mix of data center to other segments, Nvidia’s Q3 guidance puts it on the brink of achieving a $200B data center run rate with Q3 data center at $48.5B DC at midpoint.
We had outlined more than a year ago in our free analysis that analyst estimates were too low, Nvidia Q1 Earnings Preview: Blackwell And The $200B Data Center stating: “These are the current estimates, yet if the analysts are correct, then the far right of the graph will end in $50B quarterly revenue. The difference between the current consensus and this much higher trajectory can be summarized in one word: Blackwell.” At the time, consensus was that we end the year with $33B in revenue.
In fact, if we grow 10% QoQ in Q4 (it’s likely to be higher), then the data center will have quarterly revenue of $53.4 billion or about 62% higher than where analyst estimates were in May of 2024. This is a substantial disconnect for the world’s most valuable company (and a profitable one for investors who track this).
The reason this matters is that if we nail the $50B quarter (very likely we do at $53.4B) then we are on track to hit a $75 billion data center quarter by the end of fiscal 2027. This assumes a 50% CAGR over 6 quarters and 10%-11% QOQ data center growth. Keep in mind, capex just grew 23% QoQ – which helps illustrate why 10% to 11% QoQ should be doable.
If we do reach a $75 billion data center quarter compared to the $47 billion quarter that was recently reported, then that’s tracking about 60% growth in the AI segment. When you layer-in that we like to buy low when we can, and the market likes to sell semiconductors stocks at cyclical lows (when it’s really the best time to buy) we may see even more upside.
Note: I elaborated on this pre-earnings in various interviews this week including Bloomberg, Schwab and Fox Business. Bloomberg, Schwab and Fox Business.
$6 Trillion Market Cap Prediction – Revisited
Nvidia has a 30 forward PS 3-year median but even if we go with something more reasonable like a 20 forward PS, then the market cap would be $6 trillion for $300 billion data center revenue.
From my perspective, it’d be better to get the stock lower if there is high confidence in reaching this market cap.
$10 Trillion Market Cap – Revisited
To get to a $10 trillion market cap at a 20 forward PS, Nvidia’s revenue would have to be $500 billion. Analysts have the company reaching this a little after the year 2033 whereas I believe it will reach $10 trillion before 2030.
The reality is that with Nvidia’s trajectory, we could see this market cap as early as 2028. That’s because we have not factored in software to the equation, which is expected to be an equal size market as hardware. The timing on software is tricky, and thus the I/O Fund has primarily preferred lower risk, yet also hypergrowth AI hardware stocks.
Nvidia essentially has to grow the data center at a 30% CAGR for two years from 2026-2028 to reach my prediction of $10 trillion market cap two years early.
That would mean analyst estimates are five years off as analysts have Nvidia reaching this revenue into the first quarter of 2034.
Note that I'm not stuck on this happening – we will shift if the market shifts. However, it leads back to why Q3 puts us right on track as the $50B data center quarter is achievable this year (the first milestone for these predictions).
What Jensen Huang said in the Earnings Call
I’d call this the total addressable market (TAM) earnings call as that was the predominant theme.
In the opening remarks, JH stated: “We see $3 trillion to $4 trillion in AI infrastructure spend in the — by the end of the decade.” This was followed up by commentary that he sizes the market at $600 billion today. The math there is 5X growth.
These are strong comments and thus it was followed up on during the Q&A where JH expanded his comments to say:
“And so over the next couple of years, you're going to — well, you asked about longer term. Over the next 5 years, we're going to scale into with Blackwell, with Rubin and follow-ons to scale into effectively a $3 trillion to $4 trillion AI infrastructure opportunity. The last couple of years, you have seen that CapEx has grown in just the top 4 CSPs by — has doubled and grown to about $600 billion. So we're in the beginning of this build-out, and the AI technology advances has really enabled AI to be able to adopt and solve problems to many different industries.”
This was followed up on by an analyst who asked for more specifics about what Nvidia’s share of that would be. Frankly, my guess would have been about 30% to 50% for Nvidia but management stated it would be closer to high 50% to 70%.
“Benjamin Alexander Reitzes
Jensen, I wanted to ask you about your $3 trillion to $4 trillion in data center infrastructure spend by the end of the decade. Previously, you talked about something in the $1 billion range, which I believe was just for compute by 2028. If you take past comments, $3 trillion to $4 trillion would imply maybe $2 billion plus in compute spend. And just wanted to know if that was right and that's what you're seeing by the end of the decade. And wondering what you think your share will be of that. Your share right now of total
infrastructure compute-wise is very high, so I wanted to see. And also if there's any bottlenecks you're concerned about like power to get to the $3 trillion to $4 trillion.
Jen-Hsun Huang
[…] United States represents about 60% of the world's compute. And over time, you would think that artificial intelligence would reflect GDP scale and growth and so — and would be, of course, accelerating GDP growth.
And so our contribution to that is a large part of the AI infrastructure. Out of a gigawatt AI factory, which can go anywhere from $50 billion to plus or minus 10%, let's say, $50 billion to $60 billion, we represent about $35 billion plus or minus of that and $35 billion out of $50 billion per gigawatt data center.”
My thoughts are that this could be “CEO math” which is often too high. For example, JH had said 1K racks were shipping per week yet the reality is this print didn’t confirm that statement was true. The CFO did state that in this earnings call, and naturally, CFOs tend to be more accurate so let’s see if the 1K rack statement comes to fruition in the next earnings report.
Regarding what was stated above, if the current number is $600B for AI infrastructure and we have close to a $200B run rate, then 33% is a good solid number to work with, especially given energy will take front and center on costs as we go along. At $3-$4 trillion, JH is basically stating Nvidia could have a $1 trillion data center segment if Nvidia sees 33% of that by the end of this decade. That would be up from a $184B data center segment today and up from a $200B data center segment in Q3.
Additional Commentary:
Management drew attention to the magnitude that demand outstrips supply: “Right now, the buzz is — I'm sure all of you know about the buzz out there. The buzz is everything sold out. H100 sold out. H200s are sold out. Large CSPs are coming out renting capacity from other CSPs. And so the AI-native start-ups are really scrambling to get capacity so that they could train their reasoning models. And so the demand is really, really high”
In terms of why Rubin can stand its own against previous generations of GPUs and continue to drive forth demand, management expanded on the advent of reasoning agentic AI:
“At the highest level of growth drivers would be the evolution, the introduction, if you will, of reasoning agentic AI. Where chatbots used to be one shot, you give it a prompt and it would generate the answer, now the AI does research. It thinks and does a plan, and it might use tools. And so it's called long thinking; and the longer it thinks, oftentimes, it produces better answers.
And the amount of computation necessary for 1 shot versus reasoning agentic AI models could be 100x, 1,000x and potentially even more as the amount of research and basically reading and comprehension that it goes off to do.”
Lastly, keep an eye on networking as it is a subtle clue that the NVL72s are starting to move the needle – above all else, we want to see these larger systems shipping as these massive SKUs are already at $28B this quarter and have a long runway to go. As you know, we are positioned to capture the AI networking tailwinds from these larger systems:
“Networking delivered record revenue of $7.3 billion, and escalating demands of AI compute clusters necessitate high efficiency and low latency networking. This represents a 46% sequential and 98% year-on-year increase with strong demand across Spectrum- X Ethernet, InfiniBand and NVLink. Our Spectrum-X enhanced Ethernet solutions provide the highest throughput and lowest latency network for Ethernet AI workloads. Spectrum-X Ethernet delivered double-digit sequential and year-over-year growth with annualized revenue exceeding $10 billion. At Hot Chips, we introduced Spectrum-XGS Ethernet, a technology design to unify disparate data centers into giga-scale AI super factories. [ CoreWeave ] is an initial adopter of the solution, which is projected to double GPU-to-GPU communication speed.”
Conclusion:
I wouldn’t be surprised if the market continues to sell this report, as the market likes to do at cyclical lows for semiconductor companies. What’s quite impressive is that Nvidia has continued to beat even at its cyclical low (defined as being in-between GPU generations) and despite the curveball with China’s H20 revenue.
Our fundamentals process favors QoQ acceleration and the 15% QoQ growth for Q3 helps put our estimates right on track. The strong networking growth spells good things to come as it indicates it’s the larger systems that are propping up the growth (Blackwell was stated to have grown 17% QoQ).
Now is the perfect time to pause and look a the medium-term and longer-term picture to answer for ourselves if the world’s most valuable company can continue to grow – or are we in a bubble? Each Member will need to decide that for themselves although my answer is “yes, for many AI stocks we are in a bubble – but for Nvidia, we are not.” Hopefully, this post-earnings writeup helps to substantiate why I continue to believe there is room in this stock.
Please note: The I/O Fund conducts research and draws conclusions for the company’s portfolio. We then share that information with our readers and offer real-time trade notifications. This is not a guarantee of a stock’s performance and it is not financial advice. Please consult your personal financial advisor before buying any stock in the companies mentioned in this analysis. Beth Kindig and the I/O Fund own shares in NVDA at the time of writing and may own stocks pictured in the charts.
Coherent has all of the right products to potentially become a sizable player in AI networking. Primarily, Coherent’s 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.
Transceiver speed has been growing with the highest data rates ranging from 100G to 200G to 400G. AI servers are driving a market for 800G data rates, which are shipping in production now, and 1.6T rates, which are initially shipping yet expected to grow in volume come 2026.
Coherent’s transceivers work with both Ethernet or InfiniBand, as well as proprietary protocols such as Nvidia’s NVLink and Nvidia’s interconnect chips NVSwitch. The company has stated that their 100ZR pluggable transceivers can upgrade old 10GBps Ethernet links with 100 GBps at the optical network edge, representing a 10X upgrade.
Coherent designs and manufactures the components, such as lasers, detectors and passive optics. Manufacturing the components (as opposed to buying them) is a strength as the company has supply chain resiliency by controlling the end-to-end process, and the manufacturing takes place primarily in the United States with some in Europe.
What COHR must navigate is that other companies also offer optical interconnects, and thus, competition is heightened in the datacom transceivers, silicon photonics, and CW/EML/VSCEL laser space. For example, Chinese companies such as InnoLight and Eoptolink compete with Coherent, larger companies like Broadcom, and then smaller companies with agile engineering like Lumentum also directly compete. In fact, Lumentum’s report was stronger than Coherent’s report this quarter in a a few key areas which likely drove COHR’s weak price action.
In terms of fundamentals, Coherent was weak in terms of the company guiding for lower sequential growth in its datacom segment than the prior two quarters. The company’s margins are also coming under pressure, which is the most important line item for a hardware company in terms of investor appetite.
With that said, I don’t think Coherent is out of the game by any means. There are some important things in the pipeline – all discussed for you below.
Quick note on Coherent Divesting the Aerospace and Defense Business:
This quarter, Coherent divested its underperforming segment Aerospace and Defense; selling the unit for $400 million. This will help the company streamline its operations and provide more cash to help pay down elevated debt. The sale will result in Coherent exiting from 10 sites and reducing employee count by 550 employees, and thus, will be accretive to EPS.
According to analyst notes, the expected headwind to revenue will be $170 million, thus growth will appear lower in the forward growth estimates provided below although the company will be growing organically at a higher rate.
Revenue Growth Decelerating to Single-Digits
Coherent reported a slight 1.1% beat to revenue estimates, reporting Q4 revenue of $1.53 billion, up 16.4% YoY. Revenue growth has decelerated more than 10 points since the start of the fiscal year.
For Q1, Coherent offered guidance for $1.46 to $1.60 billion in revenue, or $1.53 billion at the midpoint, excluding ~$0.02 billion in revenue related to its Aerospace & Defense unit that it expects to occur after the sale is closed.
This compares to consensus estimates for $1.55 billion. On a YoY basis, this points to YoY growth of 13.3% at midpoint, with growth expected to decelerate to the mid single-digit levels in both Q2 and Q3.
For fiscal 2025, Coherent reported 23.4% YoY growth in revenue to $5.81 billion, driven by a 51% increase in data center and communications revenue, offset by a (2%) decline in industrial and other revenue. This is discussed further in detail below.
For fiscal 2026, Coherent is expected to report just 8.4% revenue growth to $6.3 billion, down from 10.0% growth to $6.37 billion at the end of July; however, the decline in growth estimates looks to stem from the Aero & Defense unit divestment. BofA had estimated the sale would create a ~$170 million headwind (implying ~$6.2 billion), yet estimates are only ~$70 million lower, suggesting other segments are offsetting some of the impact.
May 2025 Investor’s Day:
A few months back, Coherent provided updated growth targets for revenue and earnings. It’s notable that Coherent is guiding for lower top line growth in the coming years yet helps to illustrate the primary growth will be on the bottom line.
Revenue growth of 23% in FY25 versus target model of 10% to 15%
Yet, operating margin expansion expected of 6.2+ points to 24% operating margin
Key Segments: Networking Growth Decelerates to 39%
Networking remains the primary driver of Coherent’s growth, with the company announcing that it commenced revenue shipments of its first 1.6T transceiver products and its first differentiated liquid-crystal optical circuit switch (OCS) platform.
Networking revenue rose 39% YoY and 5% QoQ to $945.2 million in Q4, with its share of revenue growing by two points sequentially to 62%. For the full year, Networking revenue rose 49% YoY to $3.42 billion, accounting for 59% of revenue.
Growth has decelerated steadily throughout the fiscal year, from 61% in Q1 to 45% growth in Q3 and now 39% in Q4. However, the primary concern is that the segment was showing a rather sharp deceleration on a QoQ basis, to the lowest sequential growth since early FY24. Coherent reported just 5% QoQ growth in networking in Q4, versus 10% in Q3.
Lasers revenue declined (2%) YoY and (4%) QoQ to $348 million, the segment’s first YoY decline in five quarters. Lasers accounted for 23% of revenue in Q4, down from 24% in Q3. For the full year, Lasers revenue grew just 3% to $1.44 billion.
Materials revenue declined (15%) YoY and was flat QoQ at $236.2 million, its sharpest decline since Q1. Of the three segments, Materials was the only to see revenue decline in FY25, down more than (6%) YoY to $953.8 million.
End Markets: Data Center Reporting Low 3% QoQ Growth
Turning to end markets, Coherent has now reorganized its reporting into just two end markets: Data center and Communications, and Industrial (which includes its previously reported Instrumentation and Electronics end markets).
The main concern is that Data center growth has slowed dramatically on a QoQ basis, from 11% last quarter to 3% in Q4, despite signs of accelerating AI systems demand (capex spend increasing from Big Tech, Blackwell shipping, Blackwell Ultra seeing initial shipments, etc)
Data center and Communications revenue rose 39% YoY and 5% QoQ in Q4 to $942 million. For the full-year, revenue increased 51% to $3.44 billion.
Within this, Data center revenue increased 38% YoY but just 3% QoQ. For FY25, Data center revenue rose 61% YoY. Similar to Networking, the major concern here is that Data center growth has slowed dramatically on a QoQ basis, from 11% last quarter to 3% in Q4, despite those signs of accelerating AI systems demand.
Management did state that “sequential growth rates can fluctuate quarter-to-quarter based on lumpiness of demand from our customers or supply or capacity related things,” but they offered little to soothe fears of rising competitive pressure within Nvidia’s supply chain. Aside from saying the see strong demand ahead, management dodged the question about when Data center’s QoQ growth would accelerate.
In a brief update on the Data center product roadmap, Coherent said it expects 1.6T transceiver volume to ramp throughout calendar 2025 with more meaningful contribution in calendar 2026, while demand continued to grow in Q4 for <1.6T data rates.
For Communications, Coherent said Q4 saw accelerated growth, up 11% QoQ and 42% YoY. For the year, Communications revenue grew 23%. Management stated that the 100G ZR product family is ramping rapidly, and they expected increasing revenue contribution through FY26 from 100G, 400G and 800G ZR/ZR+ transceivers.
Industrial revenue declined (8%) YoY and (2%) QoQ to $587 million, while revenue for the full-year declined (2%) YoY to $2.37 billion. Coherent said that above-market growth in industrial lasers and services was offset by a decline in silicon carbide, consistent with softer end market demand from autos. Management said that silicon carbide has stabilized and is not expected to be a headwind in FY26.
Margins Expand YoY in FY25
Despite gross margin expanding in Q4, GAAP operating margin shrunk, pressuring GAAP EPS; however, adjusted operating margin met management’s guidance for the quarter. For the full-year, Coherent delivered expansion for gross and operating margins.
Q4 GAAP gross margin was 35.7%, up 2.8 points YoY and half a point QoQ. Adjusted gross margin was 38.1%, slightly above the midpoint of guidance for 37-39%, up 2.3 points YoY but down 0.4 points QoQ.
Q4 GAAP operating margin was 0.4%, down 4.4 points YoY and QoQ; as a result of the pending divestment, Coherent recorded $85 million in asset impairment charges, impacting the margin. Adjusted operating margin was 18%, up 2.6 points YoY but down 0.6 points QoQ.
Q4 GAAP net margin was (6.3%), down 2.6 points YoY and down 7.3 points QoQ. Adjusted net margin was 12.6%, up more than 4 points YoY and nearly 1 point QoQ.
For Q1 FY26, management offered guidance for adjusted gross margin and adjusted operating expenses:
Adjusted gross margin was guided between 37.5% to 39.5%, up 1.8 points YoY and 0.4 points QoQ at midpoint.
Adjusted operating expenses were guided between $290-310 million, implying adjusted operating margin at 18.9%, up 2.8 points YoY and 0.9 points QoQ.
For FY25:
GAAP gross margin expanded 4.3 points YoY to 35.2%, while adjusted operating margin expanded 3.6 points YoY to 37.9%.
GAAP operating margin increased 3 points YoY to 5.0%, while adjusted operating margin increased 4.7 points YoY to 17.8%.
GAAP net margin expanded 4.1 points YoY to 0.8%, while adjusted net margin expanded 3.8 points to 11.9%.
EPS Beat in Q4, Guidance In-Line Q1
Coherent reported an 8.7% adjusted EPS beat in Q4, though offered guidance for Q1 in line with consensus estimates.
Q4 adjusted EPS was $1.00, ahead of estimates for $0.92 and representing YoY growth of nearly 64%. For Q1, Coherent guided for $0.93 to $1.13 in adjusted EPS, in line with the $1.03 estimate at midpoint and representing a deceleration to ~39% YoY growth. The deceleration is stemming from minimal expansion in adjusted margins in recent quarters combined with the top-line deceleration.
For FY25, Coherent reported 192% YoY growth to $3.53 in adjusted EPS. FY26 is estimated to see growth moderate to 30% YoY to $4.59, while management noted that the Aero & Defense divestment is expected to be accretive to EPS.
Cash and Balance Sheet
Cash flows moderated and cash flow margins shrunk to the lowest levels in the past six quarters. Coherent also noted that proceeds from the divestment will be used to pay down debt.
Operating cash flow was $130.3 million in Q4, down approximately (20%) YoY and QoQ. OCF margin was 8.5%, down nearly 4 points YoY and the first single-digit margin in the last five quarters.
For FY25, operating cash flow rose 16% YoY to $633.6 million, for a 10.9% margin, down 0.7 points YoY due to the softer Q4.
Free cash flow was ($1 million) in Q4, for a (0.1%) margin, down nearly 5 points YoY. It also was the first quarter with negative FCF since Q2 FY24.
For FY25, free cash flow was $192.8 million, down (3%) YoY. FCF margin was 3.3%, down nearly 1 point YoY.
Cash and equivalents were $909.2 million, while debt was $3.69 billion.
Inventories rose 3.6% QoQ to $1.44 billion.
Earnings Call Q&A:
800G and 1.6T Shipping, yet Data Center reporting declining QoQ growth
As discussed in our previous writeup on Coherent, the company supplies EML Lasers, VSCEL Lasers and CW Lasers for silicon photonics. While 100G per lane for 400G and 800G optical transceivers is what is supporting the growth now, it’s expected that 200G per lane and even 400G per lane for 1.6T optical transceivers is what will drive growth in the coming quarters.
Management offered the following update in terms of 800G ramping now and 1.6T ramping i the coming quarters:
“Okay. On the first part of the discussion, the way — I think the way to think about it is if you start with the 800-gig ramp, the 800-gig is obviously growing this calendar year versus prior. We expect 800-gig to grow again next calendar year, and that's ramping very quickly. And then on top of that, 1.6T, we believe, starts to ramp on top of that 800-gig ramp. We saw initial revenue in the prior quarter. We expect that revenue to grow over the coming quarters.”
In the May investor’s day, the company provided the following timeline for the ramp of the new data rates, showing Coherent has a healthy pipeline over the next few years.
However, despite these updates – the market is nervous that Coherent is not able to compete given data center sequential growth is declining QoQ. There was a pointed question on the call on the unusual QoQ decline Coherent is expecting (lumpiness). Although the answer from the CEO does not directly address the timing issues, the concern is significant enough to quote the exchange in full:
Vivek Arya, BofA:
So the first one, Jim, I realized this is a little bit more short-term oriented. But when I look at your data center and communications segment, sequential growth rate has gone from 9% in March to 5% in June, and I think your September implied is probably at or somewhat below this number, even though you're starting to ramp 1.6T and OCS. So what is the right way to interpret, right, this kind of somewhat slowdown because when I look at the deployment of AI clusters, they seem to be accelerating in the back half and one of your closest peers guided to double-digit sequential growth. So how would you address that pushback and do you think the sequential growth rates can start to reaccelerate at some point?
James Robert Anderson, CEO
Yes. Thanks, Vivek. On a quarter-to-quarter basis, the sequential growth rates can fluctuate quarter-to-quarter based on lumpiness of demand from our customers or supply or capacity related things, so. But I think if you look over the full year of fiscal '25, I'm quite pleased with our growth in data center for full year. We saw over 60% growth and even faster growth in the higher speed data rates. And we believe over that fiscal '25, we gained share over that fiscal '25. We feel good about fiscal '25 results. And as I said, looking forward into the current fiscal year, again, we see very strong demand ahead of us, and a number of different growth vectors, 800-gig, 1.6T. We talked about OCS as well — as well as seeing very strong demand in our DCI segment. So we feel good about the growth ahead of us. And certainly, making sure that we've got all the capacity in place to meet that — those demand signals.
InP Capacity Tripled plus 6-inch production line offers additional capacity
EMLs were traditionally used by telecom customers, yet became attractive for AI servers due to meeting the 200G per second speeds necessary for 1.6T optical modules to support AI models. These are called single mode optics, made of Indium Phosphide, which has been used instead of silicon for long-haul networking due to being a superior choice for optical functions, such as enabling the laser, modulator, photodetector and amplifier.
InP is more expensive at the component level as four EMLs are needed compared to two lower-cost CW lasers for silicon photonics modules, yet this difference at the component level can be made up for in data centers as InP reduces power consumption.
Setting aside the data center segment lumpiness, one reason Coherent may not be down for the count (time will tell) is they have recently tripled their InP capacity and are rolling-out a 6-inch InP production line for yet another significant increase in capacity.
According to Coherent, this will be the world’s first 6-inch wafers for InP. Inevitably, there were questions on the call about how quickly Coherent could produce more EML and CW lasers plus Co-Packaged Optics (CPOs) with this new production line:
“This is a big benefit to us in 2 ways, both capacity, obviously, on a larger wafer size we get a significant increase in capacity. But also, we expect a significant cost structure advantage. And so as we fully ramp that capacity both the ability to just drive higher volume, but lower cost structure is a big advantage for that. So we're really excited about that and looking forward to that ramp.”
Apple Partnership on VSCEL Lasers:
Apple and Coherent have partnered in the past, yet there was a new partnership announced recently for VSCEL lasers to be used in iPhones and iPads. Per the opening remarks:
“As mentioned in a recent Apple announcement regarding their American manufacturing program, we've entered into a new multiyear agreement with Apple for a new generation of VCSEL products that support Apple's iPhone and iPad products. We expect revenue from this expanded partnership with Apple to begin in the second half of calendar '26.”
Analysts are penciling in 2027 as the year where Coherent could see a more meaningful boost in revenue although despite this takeaway, management did repeat impact would be seen H2 CY2026:
“Yes, we feel really good about that expansion. I would describe it as an expansion of the partnership. It's a new generation of VCSELs that go into Apple iPad and iPhones and we do see that as an increase in revenue that will start to have impact to our revenue in the second half of next calendar year, so second half of '26”
Additional Key Points:
There are many moving pieces with Coherent and the following are also notable points from the earnings call:
Optical circuit switching will increase their TAM by $2 billion. OCS was covered in the past here. Coherent has a more advanced approach to OCS through liquid crystal technology versus the more mechanical MEMS technology that competitors offer.
Communications revenue is particularly resilient with 11% QoQ growth (outpacing DC when you separate the two). Driving this QoQ growth is DCIs (data center interconnects). Although recognized outside of the data center segment as part of telecom, the demand is driven by AI as the long-distance data transmissions were traditionally used for telecom purposes, yet and can range up to hundreds of kilometers and are now seeing demand for data center buildouts.
USA footprint: Coherent has 20 U.S. manufacturing locations in 13 states, an advantage should trade wars heat up.
Conclusion:
Coherent is in all the right place, but they are certainly not alone. The market and BofA analyst are correct to question why Coherent’s data center growth is declining QoQ especially given their competitor Lumentum had a solid earnings report. Lumentum saw 16% QoQ growth in Cloud & Networking in its Q4 and guided for ~10% QoQ in Q1.
Ultimately, we had trimmed Coherent quite a bit prior to the report simply because we felt Credo and Astera Lab deserved higher allocations. After the report, I feel the same – which is that Coherent has some work to do to not only secure a spot in our portfolio but to compete with the strong AI networking stocks we already own.
On our Discovery tier, we discuss if stocks we don't own such as Lumentum are a buy or not, plus other Nvidia Blackwell beneficiaries and AI data center energy plays. Portfolio Manager, Knox Ridley, also recently held a Discovery webinar discussing setups for new stocks. Sign up now to read more of our cutting-edge research and to view the 1-hour webinar.
Current Pro and Advanced Members: To subscribe to Discovery with 30% off, please click here to email us or email premium@io-fund.com and mention code DISCOVERY30.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.
Damien Robbins, Equity Analyst at I/O Fund contributed to this analysis.
Please note: The I/O Fund conducts research and draws conclusions for the company’s portfolio. We then share that information with our readers and offer real-time trade notifications. This is not a guarantee of a stock’s performance and it is not financial advice. Please consult your personal financial advisor before buying any stock in the companies mentioned in this analysis. Beth Kindig and the I/O Fund do not own shares in COHR at the time of writing and may own stocks pictured in the charts.
Cloudflare offered a solid earnings report, yet the valuation is extraordinarily high, thus, the company really needed to report a blowout to justify the valuation (which did not happen). It’s good to see Cloudflare report steady growth, however, and its key metrics supporting sustained growth as we patiently wait for Cloudflare to lead in AI inference at the edge.
There is a quiet strength in Cloudflare’s fundamentals and key metrics. For example, Cloudflare passed a $2B run rate for the first time, signed their first $100M deal, dollar-based net retention (DBNRR) seems to have bottomed along with a slight 1.3% acceleration in revenue. Regarding the bottom line, Cloudflare is certainly stronger than many cloud peers yet tends to walk a razor’s edge due to capex.
Regarding AI inference, for the Workers Platform known as Act 3, management connected some important dots on the earnings call as to why agentic AI will drive forward the massive inference trend. The I/O Fund team recently dug up a stat inference is expected to account for 60% to 70% of AI workloads by 2030. In particular, Cloudflare emphasizes their position is what will help the company win this market: “The fact that we sit in front of so much of the web and that more than half of our dynamic traffic is already between APIs means that we are strategically positioned to deliver the agentic web of the future.”
Cloudflare also introduced Act 4 – a new product that will help AI search engines connect with (and potentially) pay publishers for using derivatives of their copyrighted works. Although the amount of demand for this and exactly how Cloudflare will monetize this new product is not clear, it is interesting management feels confident enough to call the new use case its fourth act.
Given Cloudflare’s valuation, the entry is probably the most important aspect of this stock right now – whereas in the medium to long-term the most important aspect is timing for the broader inference market.
Cloudflare Raises FY25 Revenue Growth to Nearly 27% After Largest Beat in Six Quarters
Cloudflare reported its largest revenue beat in the last six quarters at 2.1% above consensus, with Q2 revenue up 27.8% YoY to $512.3 million. This also marked a slight 1.3 point acceleration on the top-line from 26.5% growth in Q1.
For Q3, Cloudflare guided for revenue of $543.5 to $544.5 million, ahead of estimates at the time for $538.9 million. This corresponds to a slight deceleration to the mid-to-high 26% YoY growth range, where Cloudflare is expected to remain through Q4. This provides no clear indication yet that the company is able to drive a sustained revenue acceleration aided by AI.
However, not even three weeks from the report, the Street is already getting more optimistic and is now expecting a stronger Q3. Consensus estimates are now above the high-end of management’s guidance at $544.9 million, essentially already pricing in stronger momentum fueling a beat for Q3 despite how early it is in the quarter.
For the full-year, Cloudflare raised its outlook to $2,113.5 million to $2,115.5 million, for YoY growth of 26.7%. This is a $22.5 million increase at midpoint from Cloudflare’s prior outlook for $2,090 million to $2,094 million for growth of 25.3%.
DBNRR Inflects 3 Points to 114%, Paying Customer and cRPO Growth Remains Strong
Cloudflare also showed strong key metrics in Q2, maintaining strong growth in paying customers and billings while DBNRR more meaningfully inflected.
Paying customers increased 27.5% YoY to 267,929 in Q2, the second quarter in a row with 27%+ growth. This is a notable improvement from 17% and 21% growth in Q1 and Q2 2024. Cloudflare stated that it added a record number of customers YoY spending over $1M and over $5M.
While Cloudflare noted that its largest customers are growing investments at the highest levels since 2022, growth in its large customer cohort ($100K+ ARR) is decelerating. Cloudflare reported 22% YoY growth to 3,712 $100K+ ARR customers in Q2, decelerating slightly from 23% in Q1 and 30% in the year ago quarter. This cohort accounts for 71% of revenue, up from 69% in Q1 and up from 67% in the year-ago quarter.
Billings increased 33% YoY to $559.2 million, a third straight quarter with growth above 30% YoY.
RPO increased 39% YoY and 6% QoQ to $1.98 billion.
Current RPO accounted for 66% of total RPO, or ~$1.30 billion, increasing 33% YoY in Q2, a four point acceleration from 29% growth in Q1.This is also a notable uplift from 26% growth in the year ago quarter.
Aided by strength in its >$1M customer cohort, which management said served as a tailwind, DBNRR more meaningfully inflected in Q2 to 114%, its highest level in more than a year.
Quick note on Pool of Funds:
We previously covered pool of funds here, explaining that pool of funds accounts are unique to the largest customers (for example, 4 of the top 10 customers are this account type) that use many products across the entire Cloudflare platform. These are considered larger platform deals that are paid on a monthly basis in a multi-year contract rather than an annual contract on one product. For some time, this shifted how DBNRR and RPO were reported since revenue is recognized as the customer consumes the service.
The current update (albeit a bit vague) is that “pool of funds deals with our largest customers represented low double digit in the second quarter, up from less than 3% a year ago. So significant progress.”
GAAP Margins Drift Lower
Gross margins drifted lower in Q2, driven by both an increase in depreciation expenses and in allocated costs from higher network traffic from paying customers. GAAP operating margins followed, moving further away from reaching break-even.
Cloudflare had an interesting comment on long-term margins, stating that it expects to remain comfortably in its 75% to 77% adjusted gross margin target despite passing on substantial savings to Workers’ customers. This suggests that upside to operating margins will be driven by expenditures, such as moderating higher sales & marketing spending, at 36% of revenue versus its target range of 27-29%, and high SBC at 24% of revenue.
GAAP gross margin was 74.9% in Q2, down nearly 3 points YoY and 1 point QoQ. Adjusted gross margin was 76.3%, down 2.7 points YoY and 0.8 points QoQ.
GAAP operating margin was (13.1%), down 4.4 points YoY and 2 points QoQ. Adjusted operating margin was 14.1%, approximately flat YoY and up 2.4 points QoQ; this was also ahead of guidance for 12.6%.
For Q3, Cloudflare guided for adjusted operating income of $75-76 million, pointing to adjusted operating margin of 13.9%, down nearly 1 point YoY and moderating slightly QoQ.
GAAP net margin was (9.8%), down 6 points YoY and 1.8 points QoQ. Adjusted net margin was 14.7%, down 2.6 points YoY but up 2.5 points QoQ.
FY25 EPS Raised Slightly
Cloudflare topped estimates in Q2 driven by the revenue beat and stronger adjusted margins, and boosted its FY25 adjusted EPS outlook as a result.
GAAP EPS was ($0.15), missing estimates for ($0.08) as GAAP margins drifted lower.
Adjusted EPS was $0.21, beating estimates for $0.18, fueled the outperformance in adjusted operating margin in the quarter.
For Q3, Cloudflare guided for $0.23 in adjusted EPS, a slight uptick sequentially, while for FY25, the company raised its forecast from $0.79-$0.80 to $0.85-$0.86. This corresponds to growth of ~14.5% YoY, up from the mid-6% range previously. Growth is expected to be much stronger in FY26 at ~30% YoY to $1.12.
Cash Flow Margins Contract, and Cloudflare Raises $2B in Convertible Debt
Cash flow margins contracted sequentially, while Cloudflare significantly bolstered its cash pile after a large convertible note issuance.
Operating cash flow was $99.8 million for a 19% margin, flat YoY but down from a 30% margin in Q1.
Free cash flow was $33.3 million for a 6% margin, down 4 points YoY and 5 points QoQ.
Network capex was 11% of revenue in Q2, down from 17% of revenue in Q2. Cloudflare stuck to its guidance for network capex to be 12-13% of revenue for the year, suggesting slight moderation in 2H.
In June, Cloudflare raised $1.97 billion in new convertible notes due 2030, raising its cash on hand to $3.96 billion while convertible notes outstanding rose to $3.26 billion.
Earnings Call Q&A:
Agentic AI & Small Language Models (SLMs) is where Cloudflare’s AI impact will become more apparent
Today, AI agents are mainly LLM-based copilots and assistants rather than truly autonomous agents. The broader vision is for agents to set goals and act without a human prompting each action. Gartner is a reasonable forecaster and is placing 2027-2028 as the time frame when 33% of enterprise software apps will include agentic AI.
While large language models will continue to provide the more complex reasoning tasks, small language models will be deployed to execute the day-to-day automation with major benefits over LLMs such as being quicker and cheaper when running thousands or millions of decisions per user. The major difference is that LLMs are massive knowledge engines whereas SLMs are run locally and offer speed.
That’s a critical distinction to make when listening to Cloudflare’s management talk about why their stock has not performed as well in the LLM phase of AI versus the upcoming SLM phase:
“We would not be today the right place for one of the really massive LLMs to run because those, in many cases, will require multiple different machines working in coordination. It is a more complicated task. But for smaller models, we're finding that Cloudflare is the best place for anyone who's building that to run that. And over time, we are investing in making our systems able to support larger and larger and larger models.”
In addition, Cloudflare represents 20% of the internet and that number is in the mid-30% when considering the top 10,000 sites. By managing a leading percentage of internet traffic, agentic AI will route through Cloudflare’s reverse proxy and CDN to fetch data and trigger APIs. The company offers Cloudflare Workers and WebAssembly for a low-latency and distributed hosting environment to run SLMs. Additionally, agentic AI will require enhanced security since autonomous actions can have a larger impact from errors (bad API calls) or security issues that are more malicious.
This is how management discussed it on the earnings call:
“I think what we feel confident, though, is that because of the fact that so much of the Internet sits behind us and inherently, those agents are going to be passing through us that we have an opportunity to help define what those rails are that the agents will ride on and take some fee from that — those transactions as we've helped facilitate them and make them faster, more reliable, more secure, give people the access to those rails.”
Act 4 Seeks to Minimize the Impact AI has on Publishers
Cloudflare recently launched a product that prevents AI bots from accessing websites without payment. It’s no secret that AI has made a dramatic impact on publishers given AI-driven search tools do not refer traffic to the sources they scrape data from. Whether it’s OpenAI and Microsoft getting sued by The New York Times or stats like this one that publishers are seeing 25% fewer referral clicks, the evidence is mounting that AI is destroying business models for both smaller, independent publishers and larger media conglomerates.
Cloudflare had some wild statistics on their call, stating “based on the data that Cloudflare has observed, it's nearly 10x harder to get traffic from Google than it was just 10 years ago.” Even more crazy, they stated “every AI company we've tracked is worse than the Google of old with some being as much as 30,000x harder to get traffic from” – likely referring to OpenAI or Anthropic.
This is one to watch as it could become one of Cloudflare’s fastest growing products given the pain point Act 4 seeks to solve.
“But we aren't building search engines anymore. We're building answer engines. And the difference between a search engine and an answer engine is a search engine directs you to that content where you can go and the content creator can monetize it. An answer engine answers without you having to leave. And so there has to be some value creation back to content creators that isn't just based on traffic.”
Cloudflare trades at a forward PS ratio of 28.5. The exuberant investor will tell you that any contribution from AI demands a higher valuation than what we’ve seen in the past, yet time will tell if that is true. We prefer to see if we can get Cloudflare and really any cloud stock under 20 forward PS. Typically, Cloudflare at a 15 forward PS is a great spot to buy.
Conclusion:
As of now, we are patiently waiting to get Cloudflare lower ahead of the inference market taking off. Remember – our thesis is about Cloudflare’s positioning, which is multi-faceted in terms of how Cloudflare can monetize its AI utility at the internet infrastructure level. As agentic AI and SLMs become a leading paradigm, Cloudflare’s network is set to become a critical player. You’re seeing some of this with Cloudflare able to seamlessly pivot toward helping publishers get paid by Big Tech, which is typically a cutthroat group. The company is also effortlessly expanding its use case for AI inference to include agentic AI and SLMs; meaning no matter where AI development takes inference, Cloudflare will be there.
Since our firm likes to be prudent, margins are not actually bulletproof as Cloudflare is not GAAP profitable and margins contracted this quarter. This line item has a red mark on our checklist, and we will continue to look at this closely in the coming earnings reports.
Damien Robbins, Equity Analyst at I/O Fund contributed to this analysis.
Please note: The I/O Fund conducts research and draws conclusions for the company’s portfolio. We then share that information with our readers and offer real-time trade notifications. This is not a guarantee of a stock’s performance and it is not financial advice. Please consult your personal financial advisor before buying any stock in the companies mentioned in this analysis. Beth Kindig and the I/O Fund do not own shares in NET at the time of writing and may own stocks pictured in the charts.
Reddit’s stock has surged 62% in one month, easily placing the company’s earnings report as one of the best to come out of the tech sector this quarter. The world’s leading forum site has only 416 million weekly active users compared to Facebook’s 2 billion yet ranks fifth behind Facebook as the most visited site in the United States. In addition, due to a few changes in how Google surfaces content with AI overviews, Reddit is now the second most visible site in the United States – ranking above Facebook for example – and the top line results show the company is reaping the rewards of being in the search giant’s good favor.
Therein lies the risk, which is that Reddit’s current growth is dependent upon how a Big Tech company graciously offers visibility on its platform – something that is entirely outside of Reddit’s control. The partnership could continue for years, or it could stop at any time.
Below, we offer a few ways to approach the opportunity that Reddit presents and what an investor should look for to determine if Reddit will continue to be an anomaly in that AI-driven search benefits the social platform (whereas it hurts traffic for most other publishers). The analysis also discusses what to watch for in terms of Reddit hitting peak growth and the catalysts that could drive the stock further.
Reddit stock surges 62% in the weeks following the Q2 2025 earnings report. Source: YChartsYCharts
Reddit’s Web Rankings are Surging; Driving Stock Returns
Reddit delivered a rather impressive Q2 on July 31st with revenue beating estimates by more than 17%. Growth translated to the bottom line as GAAP net margin expanded to 18%. Driving much of these results is the new improved web rankings that Reddit has seen over the past year due to Google’s AI Overviews.
Over the last two years, Reddit has seen an explosion in SEO visibility on Google, with data from Sistrix placing growth from July 2023 to April 2024 at a whopping 1,328%. This moved Reddit from 85th most visible site to the 7th most visible.
Now, as of August 2025, Reddit has moved to 2nd most visible site in the US, per Sistrix, behind Wikipedia, and ahead of popular sites such as Facebook in 7th, Amazon in 4th, and YouTube in 3rd. This major improvement in SEO ranking may be a potential contributor to Reddit’s accelerating growth over the past five quarters – yet as stated, the surge in growth is out of Reddit’s control and relies on Google SEO placement, which could change at anytime.
In terms of user engagement, Reddit notched 4 billion visits as of July 2025, according to data from Similarweb, compared to 12.1 billion for Facebook, 6.6 billion for Instagram, and 4.5 billion for X. Users visited an average of 4.83 pages per visit with an average visit duration of 6 minutes, compared to 12.57 pages per visit and an average duration of more than 10 minutes for Facebook. Similarweb places Reddit as the fifth-most visited site in the US, behind Facebook in fourth place.
Advertising Revenue Rises 84% YoY
Behind the substantial revenue beat in Q2 was 84% growth in advertising revenue to $465 million. This marked a sharp 23 point acceleration from 61% growth in Q1. Sequentially, advertising revenue grew almost 30%, with growth of more $106 million QoQ, outpacing Q4 2024’s $79 million sequential increase. This is impressive considering Q4 is typically the strongest seasonally for the company.
Reddit’s stock price surges after reporting advertising growth of 84% YoY
Reddit said that the majority of this advertising growth was driven by existing advertisers deepening investments on the platform, though it did see 50% YoY growth in active advertisers as it continued to acquire new advertisers. Additionally, performance ads and brand ads both increased more than 80% YoY, reflecting strong engagement from advertisers on the platform.
Reddit also highlighted ways it is driving higher return on ad spend for advertisers and higher click-through rates, both key factors in increasing ad spend to maintain a robust ad revenue flywheel. Management said that Dynamic Product Ads for shopping were made generally available in Q2, and advertisers were consistently achieving 2x higher ROAS on average versus standard campaigns. Reddit also said that Conversation Summary Add-Ons, which integrate positive user conversations directly into product or brand ads, were delivering more than 10% higher click-through rates versus standard image ads.
Reddit Monetizes through Ads, yet is Expanding its Monetization Methods
Similar to most large audiences online, Reddit monetizes through ads. While ads are its main revenue stream, the company has recently branched out into premium subscriptions and data licensing deals.
Reddit’s recent product updates target all three of its stated growth levers: core product, search and international growth.
In core product, Reddit is working to boost engagement and platform stickiness by lowering barriers to log-in, using AI to improve personalization of homescreens and community discovery, and improve user onboarding. Bringing more logged-in users to the platform is especially important considering logged-in users contribute a higher ARPU than logged-out, yet this cohort accounts for under 45% of daily active unique users (DAUq).
Reddit Search is another product feature that now has 70 million weekly active unique users (WAUq) as of Q2, or a nearly 17% attach rate to its global WAUq of 416.4 million. Within this is Reddit’s AI-powered search engine Reddit Answers, which saw WAUq rise 500% sequentially in Q2, from 1 million to 6 million, or nearly 10% attach of Search users.
Answers are separate from Search in Q1, although Reddit is now working to unify the two centrally in the same search-box, assisting in queries from new users, existing users, and users routed to Reddit via external search engines.
Query growth has been very strong, with Reddit noting that average query volume rose more than 10X sequentially from Q1 to Q2.
Reddit Seeks International Expansion: Double-edge sword
Global expansion for Answers is continuing, with the feature only available to US users and some international markets including the UK, India, Australia and Canada. While International user growth is outpacing US growth by a large degree, Reddit expanded the rollout of its AI-powered automatic Machine Translation feature from 13 languages in Q1 to 23 in Q2.
Available across 35+ countries worldwide, Machine Translation removes barriers between users in different geographies, by allowing them to “post and comment in their preferred language, which will be auto-translated into the community’s native language.”
For example, Reddit says that users in Spain who typically use Reddit in Spanish could seamlessly contribute to French communities and threads. By lowering barriers to communicate between different geographies and languages, Reddit is opening the door to further expand its 235 million international user base (for comparison, Meta’s international active users are 11x higher).
This is a double-edged sword, however, as the United States region presents a hidden risk and could create a headwind to growth as user growth stalls, considering the region contributes the bulk of revenue.
Reddit’s WAUq Shows Stronger International Growth, US Plateau
International weekly active unique users (WAUq) growth remains quite strong, with Q2 being its sixth consecutive quarter of >30% growth. Meanawhile, United States WAUq is plateauing, with growth decelerating 60 points over the past year and 10 points in the past quarter to just 8% in Q2.
Global WAUq rose 22% YoY and 4% QoQ to 416.4 million.
US WAUq rose just 8% YoY and 1.5% QoQ to 181 million. In terms of user additions, YoY growth was just 13.5 million.
International WAUq rose 35% YoY and nearly 6% QoQ to 235.4 million, or user additions of more than 60 million YoY.
Why Slowing United States User Growth is not Ideal
The plateau in US WAUq with minimal growth since Q3 2024 raises the risk that Reddit’s US penetration is approaching a peak and already maturing. With estimated US internet users of 322 million, Reddit’s penetration would be around 56%, versus Meta at 84% with 272 million MAUs in the US as of Q4 2023.
Should Reddit find it difficult to boost its US penetration towards 200 million and above, or if US user growth does plateau here, future revenue gains will hinge on two key monetization drivers: impressions growth or ARPU growth.
Driving impressions growth higher would rely on increasing ad loads either via higher platform engagement or content consumption (higher time spent or pages visited), or greater ad density. Ad density is a rather fine line to walk without oversaturating the platform with ads to the detriment of the user experience. ARPU growth would stem from improving ad pricing from better targeting or higher-value, higher-converting formats (increasing ROAS for advertisers), which Reddit is currently working on.
This also ties into international growth – International monetizes at just over one-fifth the rate of US, yet now accounts for almost 57% of WAUq, up from 51% a year ago. If US plateaus and International continues to grow much quicker, say at >20% YoY, this presents a real headwind to global ARPU since the region monetizes at a much lower rate yet accounts for an increasingly larger share of users.
The One Metric that Reddit Investors Should Watch Closely
Reddit was the worst YTD performer against peers in June when fears were rising around traffic hits from Google’s AI Overviews, with shares down more than (30%) YTD. Barely two months later, shares have doubled, now making Reddit the best performer with its YTD return of 37%.
We believe one key metric – above all other numbers – will help determine if Reddit can sustain darling status in 2025 and beyond. Given the meteoric rise this past month with Reddit’s stock seeing gains of 62% in one month, the information below is not to be missed.
Sign up below to access the following information:
The one key metric that provides an important hint as to whether Reddit can sustain its stock trajectory.
Breaking down in granular detail why The Street is buying the stock; information we haven’t touched on yet in this analysis
Why top line growth is important but may not be what ultimately drives the stock’s performance – and what will drive the stock further instead
Reddit’s ARPU Accelerates to Record High
The one metric that suggests Reddit’s run may not be over is the impressive acceleration seen in ARPU this past quarter.
This past quarter marked the highest sequential growth in ARPU in more than three years at 25%, outpacing even Q4 24’s 18% growth. This could spell good things for the upcoming Q4, which is a seasonally high quarter for ad companies.
Global ARPU was $4.53 in Q2, accelerating 24 points to 47% YoY, driven primarily by growth in ad impressions and to a lesser extent, growth in ad prices. Ad pricing still did provide a “nice tailwind” in Q2, per Reddit, accelerating a bit further in Q2 after beginning to accelerate in Q1. Unlike Meta, Reddit does not break out impressions or pricing growth, so it’s impossible to see the pace at which the pricing metrics are growing.
Management believes global ARPU is “still low on an absolute basis and remains an opportunity” for long-term improvement – for example, Meta’s global ARPU is around 3x of Reddit’s at $13.65 as of Q2, and though Meta hasn’t updated regional metrics since the end of 2023, it’s possible that US ARPU is 10x that of Reddit’s.
US ARPU jumped to $7.87, accelerating 28 points sequentially to 59% YoY in Q2. This is a remarkable achievement, considering US ARPU had struggled to grow in early 2024, with the year ago quarter seeing ARPU decline (5%) YoY.
International ARPU rose 40% YoY to $1.73, also an 18 point sequential acceleration from 22% growth in Q1. While the region monetizes at a much lower rate than the US, similar to Meta, growth is lagging the US, presenting a headwind to global ARPU moving forward as international now begins to overtake the US in terms of active user share.
Strong ARPU drives Revenue Acceleration, but may peak in Q2
Reddit reported 77.7% growth in revenue in Q2 to $499.6 million, well ahead of estimates for just $426 million in the quarter – to put this in perspective, Reddit’s advertising revenue of $465 million alone would have beat estimates by more than 9%. AI data licensing and other revenue remains in the backseat, growing just 24% YoY to $34.8 million, or 7% of revenue.
Q2 marked Reddit’s fastest growth since the start of 2022, and a significant improvement over the past two years from just 12% growth at the start of 2023. What’s even more impressive is that Reddit delivered this 77.7% growth on top of a rather difficult 53.6% comp, yet this may shape up to be the peak growth quarter for the year as comps get tougher.
For Q3, Reddit guided for $535 to $545 million in revenue, or 55% at midpoint, against a 67.9% comp from Q3 2024. This was a strong outlook, more than 14% above consensus estimates for $473 million for the quarter. For Q4, analysts are currently expecting growth of 46.8% YoY to nearly $628 million.
It’s not all smooth sailing for Reddit, as changes to Google’s algorithm and AI overviews have provided some headwinds. Management was straightforward in Q1 in saying that they do “expect some bumps along the way from Google because we've already seen a few this year. This is expected in any year, but given that the search ecosystem is under heavy construction, the near term could be more bumpy than usual.” Reddit did state that Google traffic varies from week to week, but that it was a headwind in Q2.
Looking ahead, Reddit is currently expected to generate $2.06 billion in revenue in 2025, up nearly 59% YoY – this is 20 points faster than estimates from six months ago at 39% YoY to $1.81 billion. For 2026, Reddit is projected to report nearly 32% growth to $2.72 billion, more than doubling its revenue in just two years (from $1.3 billion in 2024).
While discussing revenue, it’s important to touch upon advertiser concentration, as Reddit is quite heavily concentrated. Reddit noted that its top ten largest customers accounted for 25% and 26% of revenue in 2024 and 2023, or $325 million and $209 million respectively.
Analysts Raise Estimates = The Street buys the stock
Reddit’s revenue and earnings revisions are strongly positive as the company’s business momentum continues to outpace estimates quarter after quarter — few stocks in the tech universe boast revenue and earnings revisions to this degree.
Over the past month, consensus EPS estimates through Q4 2026, or the next six quarters, have been revised 23% to 66% higher; over the past six months, estimates have moved 20% to 57% higher, as margins strengthen. For example, Q2 2026 has seen its estimate move from $0.42 to $0.70 over the past month, and Q3 2026 from $0.55 to $0.83. This now projects three consecutive quarters of triple-digit YoY growth followed by three consecutive quarters of >50% growth.
For revenue, Reddit is now expected to see six consecutive quarters of >30% growth, with strong revisions to estimates through mid to late 2026. Q2 2026’s estimate has been revised nearly 21% higher over the past month from $545 million to $659 million, while Q3’s revenue has risen from $601 million to $711 million.
On an annual basis, Reddit’s FY26 EPS has been revised 29% higher over the past month from $2.35 to $3.02, and revenue 14% higher from $2.39 billion to $2.72 billion.
Reddit’s Quality Fundamentals
In addition to the rapid rise in analyst estimates, which provides immediate room in the stock’s valuation, Reddit offers quality fundamentals (as many ad-based models do).
GAAP Net Margin at 18% Despite Rising Costs, High SBC
While the acceleration in advertising revenue and strong ARPU growth is certainly impressive, Reddit’s margin profile is arguably more impressive this quarter. The margin expansion is especially impressive considering SBC is quite high, and costs are rising at the highest pace in more than a year.
Reddit reported its fourth consecutive quarter with a >90% gross margin in Q2, while operating and net margin also marked their fourth consecutive quarter in positive territory.
Gross margin was 90/8% in Q2, up 1.3 points YoY and marginally higher QoQ.
Operating margin was 13.6%, up nearly 25 points YoY and 12.6 points QoQ. Notably, this also exceeded Q4 2024’s operating margin of 12.4%.
Net margin was 17.9%, up more than 21 points YoY and 11.2 points QoQ. Net margin is higher than operating margin as Reddit benefits from interest income on its $2 billion in cash on hand.
What’s remarkable here is that Reddit delivered this level of improvement in operating and net margin despite operating expenses rising nearly 37% YoY in Q2, and with SBC being quite elevated at 18% of revenue. This also implies that as Reddit matures and scales revenue into the billions in the long run, there is a pathway to potentially double net margin should expenses growth normalize back to the teens and SBC also normalize to a much lower percentage of revenue.
EPS and Adjusted EBITDA
Fueled by the large top-line beat and strength in margins down the line, Reddit delivered an outstanding 138% EPS beat in Q2, reporting $0.45 per share versus estimates for $0.19.
Looking ahead, EPS is expected to grow triple digits in both Q3 and Q4, with current estimates pointing to 209% YoY growth to $0.49 in Q3 and 111% growth to $0.76 in Q4. This implies analysts are placing Reddit’s net margin at approximately 25% by Q4, up 7 points, assuming minimal share dilution.
For the full year, Reddit is expected to report EPS of $1.86, with the majority of this coming in the next two quarters. For 2026, EPS is projected to grow nearly 63% YoY to $3.02, outpacing revenue growth by a factor of 2x as Reddit benefits from increasing operating leverage.
Adjusted EBITDA was $166.7 million in Q2 for a 33.4% margin, up more than 19 points YoY and 4 points QoQ. However, unlike operating and net margins, adjusted EBITDA margin did not surpass Q4’s level. For Q3, Reddit guided for $185 to $195 million in adjusted EBITDA, or a 35.2% margin at midpoint.
Cash Flows and Balance Sheet
Cash flows continue to be strong, with Q2 the fourth consecutive quarter with cash flow margins above 20%. Reddit’s balance sheet is very healthy as well, with no debt and more than $2 billion in cash and marketable securities.
Operating cash flow was $111.3 million in Q2, up 292% YoY. OCF margin was 22.3%, down 10 points QoQ but up more than 12 points YoY.
Due to limited capex expenditures, free cash flow and operating cash flow are correlated nearly 1:1. Free cash flow was $110.8 million in Q2 for a 22.2% margin.
Cash and marketable securities have increased more than 20% over the past year to $2.06 billion.
One more catalyst to mention: Reddit Seeks AI Data Licensing Deals:
While advertising revenue is firmly in the drivers seat for growth, accounting for 93% of revenue in Q2, Reddit does have some opportunities from AI data licensing. Ahead of its IPO in early 2024, Reddit stated that it had signed contracts worth $203 million over a two to three-year period, letting companies train AI models on its >1 billion cumulative posts and >16 billion comments.
In February 2024, it was revealed that Google was behind one of the deals, worth an estimated $60 million per year; shortly after this was announced, the FTC opened an inquiry into Reddit’s AI data licensing plans. In March 2024, Reddit signed another deal with PR firm Cision, and in May, Reddit signed a deal with OpenAI for data licensing, with OpenAI to bring ‘enhanced’ Reddit content to ChatGPT and become an ad partner on the site.
However, the AI data licensing side has seen its fair share of controversy, with Reddit recently suing OpenAI competitor Anthropic in June, alleging it was training its AI models by scraping Reddit’s data without consent. Additionally, earlier this week, Reddit began blocking internet archive the Wayback Machine from accessing its site, in order to prevent unnamed AI firms from using the archive to scrape data for free.
Valuation Far Above Peers
Reddit was the worst YTD performer against peers in June when fears were rising around traffic hits from Google’s AI Overviews, with shares down more than (30%) YTD. Barely two months later, shares have doubled, now making Reddit the best performer with its YTD return of 37%.
Interestingly, despite underperforming the sector through all of Q2, Reddit has traded at a premium on a forward PS basis the entire year, and has returned to its prior peak and frothy level at >20x forward PS. Reddit now trades at double Meta’s valuation at 10.1x forward PS, yet is admittedly a much higher growth stock. When compared to Applovin, another high growth ad model, Reddit trades similarly with APP at 24 forward PS.
On the bottom line, Reddit trades at a forward PE ratio of 60 compared to Applovin at 40 and Meta at 20 forward PE ratio. EV to forward EBTIDA is similar with Reddit trading at a premium of 42 versus App at 30 and Meta at 15.
Conclusion
Reddit is firing on all cylinders with a huge beat and raise in Q2, driven by a rapid acceleration in advertising revenue and ARPU. As of now, Q3’s guide, while nearly 15% above estimates, still implies Q2 will be the peak growth quarter for the year at 78% YoY before decelerating by >20 points into year-end. Despite topline growth being more subdued according to current analyst estimates, the company has triple-digit growth incoming on the bottom line for the next three quarters – any beat here will help the valuation.
The company also delivered GAAP operating and net margin expansion to new highs at 13.6% and 17.9%, even with costs rising in the high-30% range YoY and elevated SBC in the high-teens percentage of revenue.
The slowing United States user growth is to be watched, leaving the primary challenge of expanding ARPU globally. So far, so good there as ARPU is hinting that something important is going on with the stock’s monetization strategy, which we’ve outlined in this report.
Special FREE 1-hour webinar on how to manage risk with Bitcoin and cryptocurrencies. Join the team that identified Bitcoin’s previous top and previous bottom with impeccable accuracy – Knox Ridley of the I/O Fund and Vincent Duchaine of Wealth Umbrella – in a special webinar held on August 21st at 4:30 pm EST. More details here
Our five-year cumulative returns of 210% would place us as #2 if we were a hedge fund and #5 if we were an ETF. Learn more here
Please note: The I/O Fund conducts research and draws conclusions for the company’s portfolio. We then share that information with our readers and offer real-time trade notifications. This is not a guarantee of a stock’s performance and it is not financial advice. Please consult your personal financial advisor before buying any stock in the companies mentioned in this analysis. Beth Kindig and the I/O Fund do not own shares in RDDT at the time of writing and may own stocks pictured in the charts.
Thank you for being a valued part of our community and for supporting our mission to deliver high-quality stock analysis at an affordable price through the Essentials plan ($99/year).
As we continue to expand our research and enhance the experience for our members, we’ve made the strategic decision to streamline our offerings. Beginning August 1, 2025, the Essentials plan will be discontinued. This change allows us to dedicate more resources to our premium tiers—Pro, Advanced, and Discovery—which will deliver even deeper insights, more frequent updates, and expanded coverage.
What this means for you:
We are extending special discounted upgrade opportunities to our premium tiers for Essentials members. Please reach out to premium@io-fund.compremium@io-fund.com for details.
Archived Essentials Content will remain accessible under the Essentials tab until November 30, 2025. After that date, all Essentials research will be migrated to the Pro Dashboard for future access.
We appreciate your continued trust and support as we grow. These changes will allow us to deliver an even higher standard of analysis and research to our members—helping you stay ahead in today’s rapidly evolving market.
Innodata boosted its full-year revenue growth outlook by five points in Q2, now forecasting more than 45% growth this year, driven by strong demand, significant new deal wins and a large pipeline geared toward the second-half of the year.
However, digging into the details reveals that Innodata’s largest customer has not expanded over the last two quarters, and remains at the $135 million annualized rate as stated in Q4. This comes despite Innodata discussing expansion opportunities with this customer and a second SOW signed in Q1, with growth not yet materializing.
The CEO offered an update in Q2 that they’ve won “several new projects” with their largest customer and have others in the pipeline:
“We recently won several new projects with our largest customer and we have others in pipeline that are not yet included in our forecast, but which we think are reasonably likely. Several of these new projects are under the second SOW we reported signing with this customer last quarter. We believe that the second SOW potentially gives us access to an even larger generative AI revenue pool with this customer.”
Innodata is a $1.5B market cap stock highly dependent on the announcement of new customer wins and expanding its projects with the largest customer. Our strategy is to watch price closely for the best risk/reward on an entry; and any entry will have a tight stop.
We’ve discussed in the past that Innodata appears to have a low valuation when compared to peers. Our previous analysis also went into the details of Innodata’s products, market positioning and can be found here: “Innodata: Early-Stage AI Data Engineering; Lumpy Growth”
Revenue up 79.4% YoY
Innodata reported a slight revenue beat in Q2, with revenue coming in 3.6% ahead of estimates to $58.4 million, up 79.4% YoY. This did mark a more than 40 point deceleration in growth sequentially, yet that was expected coming into the quarter.
However, there have been some shifts in forward estimates for Q3 and Q4. For Q3, revenue estimates have been revised nearly $2 million lower since June, now projecting revenue of $59.8 million for growth of just 14.5% YoY. Q4 revenue has been revised more than $5.5 million higher to $70.9 million, for growth of almost 20% YoY. This suggest Q3 is now expected to see be a bottom of sorts before rebounding and reaccelerating into 2026.
For the full-year, Innodata raised its revenue growth guidance from 40% to >45%, now implying revenue of >$247 million, up from $239 million as of Q1. It is important to keep in mind the fluid nature of Innodata’s business, and that any new contractual agreements or expansions could have a large and/or immediate impact on revenue.
Management stated that the guidance hike was driven by “significant new deals that have been finalized since our last call as well as several deals that we believe are highly likely to close in the near term.” Any of these new deals could easily boost growth depending on how quickly they scale.
One key point that could be behind the post-earnings sell-off was some comments made by management around customer engagement and deal sizes. CEO Jack Abuhoff said that with one Big Tech customer, Innodata was “recently awarded a number of significant engagements… enabling us to forecast $10 million of revenue from this customer in the second half of this year,” up from $0.2 million over the last four quarters. Given the size of the revenue guide raise at just $8 million implied, this comment raises some questions about growth in 2H from other core customers.
Largest Customer Accounts for 58% of Revenue
Innodata provided only a brief update on customer progress this quarter, one with its largest customer and the other being the $10 million revenue opportunity discussed above.
With its largest customer, Innodata said that it won “several new projects,” with some of these being under the second scope-of-work (SOW) signed last quarter. Management added that there are other projects in the pipeline with the customer that are not yet included in its forecast, but are “reasonably likely” to be signed in the future.
Innodata’s largest customer accounted for 58% of revenue in Q2 and 59% of revenue for 1H, implying contributions of ~$33.9 million in Q2 and $68.9 million in 1H. This represents no change from Q4 (ie. no expansion) when it was stated the customer was at a $135 million annualized run rate, or ~$34 million quarterly.
Innodata also had no other >10% customers, implying the remaining customers are not spending more than $5.8 million per quarter with the company. The forecast for $10 million in 2H from another customer would likely make said customer Innodata’s second largest and a second >10% customer.
AI Segment Grew 99% YoY on Top of Tough Comps
DDS remains the key driver of Innodata’s growth due to its role in handling AI data preparation, labeling and annotation, AI training and related services.
DDS revenue grew 99% YoY to $50.58 million, though this was slightly down sequentially from $50.83 million in Q1. Although the segment’s growth barely dropped out of the triple-digit range, what’s impressive is that revenue still practically doubled YoY against a 93% growth comp.
Synodex revenue rose 4% YoY to $2.06 million, slowing from nearly 8% growth in Q1.
Agility revenue rose 11.5% YoY to $5.75 million, flat with growth from Q1.
Margins and Adjusted EBITDA expand significantly YoY
Innodata’s margins have significantly expanded YoY, which helps set Innodata apart from stocks in the $1B or $2B market cap range (Innodata is at $1.5B market cap). There was a marginal sequential decline QoQ yet not enough to matter in terms of the overall improvement seen from strong margin and adjusted EBITDA expansion over the past few quarters.
For example, GAAP operating margin was up 14 points YoY from 1.3% to 15.3% in the current quarter.
Q2 GAAP gross margin was 39.4% in Q2, down nearly half a point sequentially but up nearly 11 points YoY. Adjusted gross margin was 42.9%, down slightly from Q1 but up more than 9.5 points YoY.
Q2 GAAP operating margin was 15.3%, up 1.1 points sequentially and up more than 14 points YoY.
Q2 GAAP net margin was 12.4%, down 1 point sequentially but up from approximately 0% in the year ago quarter.
Turning to adjusted EBITDA — Innodata saw a slight sequential improvement in adjusted EBITDA margin in Q2, though on a YoY basis, adjusted EBITDA margin expanded more than 14 points. Innodata had guided for YoY growth in adjusted EBITDA with no further clarity, and has currently reported nearly $26 million for 1H ’25 versus $34.6 million for all of FY24.
Consolidated adjusted EBITDA margin was 22.7%, up from 21.8% in Q1 and 8.6% in the year ago quarter.
DDS adjusted EBITDA margin was 24.2%, up from 22.7% in Q1 and just 5% in the year ago quarter. Tracking DDS’ adjusted EBITDA is important considering the segment accounts for more than 92% of consolidated adjusted EBITDA.
Synodex adjusted EBITDA margin was 22.3%, up from 20.8% in Q1 but down from 26.3% in the year ago quarter
Agility adjusted EBITDA margin was 9.7%, faring the worst out of the three segments as this was down 4 points from Q1 and down nearly 10 points YoY.
EPS Beat by 80% yet EPS expected to decelerate in coming quarters
Innodata handily beat on EPS in Q2, reporting $0.20 in GAAP EPS versus consensus estimates for $0.11. Looking ahead, Q3 EPS estimates are following revenue in moving slightly lower, while Q4 estimates have moved slightly higher. Since our last report, Innodata: Early-Stage AI Data Engineering; Lumpy Growth, here’s how estimates have changed:
Q3 EPS has been revised $0.03 lower, from $0.17 to $0.14, for a YoY decline of more than (73%), though as a reminder the year-ago comp is technically inflated from a $5.9 million income tax benefit.
Q4 EPS has been revised $0.02 higher, from $0.19 to $0.21, for a YoY decline of (31.2%).
For FY25, Innodata is expected to report EPS of $0.76, down (14.6%) YoY before rebounding to 34.5% growth in FY26 to $1.02. To note, FY26’s EPS estimate is still unchanged since June’s writeup.
Cash Flow Declines 60% QoQ
Another blemish in Q2’s report were cash flows, which declined more than (60%) sequentially. Innodata’s balance sheet remains healthy, and the company still has the entirety of its $30 million credit line available.
Operating cash flow was $4.23 million, down (61%) QoQ. OCF margin was 7.3%, down more than 11 points QoQ. For 1H, operating cash flow was $15.1 million, up more than 139% YoY.
Free cash flow was $2.5 million, down more than (70%) QoQ. Free cash flow margin was 4.3%, down more than 10 points QoQ. For 1H, free cash flow was $11.0 million, up nearly 5x YoY.
Cash and equivalents totaled $59.8 million, and debt remained zero.
Deferred revenue was $6.5 million, down from $8.0 million in Q1.
Earnings Call Q&A
There were two topics on the earnings call Q&A session worth highlighting. The first was questions about Scale AI and if Innodata will see more customers onboard as a result of Meta’s large investment of $14.3 billion. The second was why agentic AI will drive more uses cases for the simulation data solutions such as what Innodata and its competitors offer.
Scale AI’s investment is a potential/speculative tailwind:
We covered why Scale AI’s investment could be a potential/speculative tailwind in our prior analysis stating: " Following Meta’s investment, it was rumored that Google, OpenAI and Tesla are looking elsewhere to avoid strengthening Meta at the cost of their proprietary data. Although it’s speculative, the exodus of major players from Scale AI could become a tailwind for Innodata. “
An analyst asked something similar on the Q2 earnings call. The CEO remained vague yet did state that something could materialize in the next couple of months:
“George Frederick Sutton, Craig-Hallum Capital Group:
Nice results. Congratulations. So I wondered if we could talk about during the quarter, your largest competitor, Scale AI was a large majority purchased by Meta. And we've had a few of the large tech companies come out and say they would no longer work with Scale AI. These ostensibely would be tech companies that you have statements of work with. So I'm just curious if you can kind of give us the after effect of that acquisition as you've seen it.
Jack S. Abuhoff, President, CEO & Director
[…] We have, in light of this stepped up that effort with certain companies and there are certain conversations that are going on and are now planned to be happening over the next couple of months that I think could be very exciting for us. I don't know that I can get into particulars much beyond that, but I'll reiterate that we do see an opportunity to accelerate our market presence.”
Agentic AI to drive forward major use cases:
Although agentic AI may still be a few years out before it’s commercially viable, the CEO pointed out the future of large language model (LLM) improvements lies in the quality of data. In order to have agentic AI that can tackle multivariant problems, training data will need to be supplemented with simulation training data.
The CEO stated the following on the call:
“We believe agent-based AI is going to serve as the cornerstone technology that unlocks the full value of large language models and generative AI for enterprises. Moreover, we believe that progress on Agentic AI is likely to soon result in a ChatGPT moment for robotics. Within the next several years, we believe Agentic AI will be served at the edge in hardware devices with which we will commonly interact in many respects in our lives. We believe the market for simulation data services and evaluation services to drive Agentic AI and robotics is likely to dwarf the market for frontier model post-training data.”
Conclusion:
We pointed out in our original analysis that Innodata has many competitors, and we provided an overview of how Innodata must find a path to compete with AI-native startups for data-as-a-service for AI training data. As proven by Scale AI, the market for data labeling and supervised learning is only going to grow in importance – but will Innodata be able to compete? That is a question this earnings report did not answer for investors. For an opportunity like this, we rely heavily on technicals.
Damien Robbins, Equity Analyst at I/O Fund contributed to this analysis.
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.
Innodata boosted its full-year revenue growth outlook by five points in Q2, now forecasting more than 45% growth this year, driven by strong demand, significant new deal wins and a large pipeline geared toward the second-half of the year.
However, digging into the details reveals that Innodata’s largest customer has not expanded over the last two quarters, and remains at the $135 million annualized rate as stated in Q4. This comes despite Innodata discussing expansion opportunities with this customer and a second SOW signed in Q1, with growth not yet materializing.
The CEO offered an update in Q2 that they’ve won “several new projects” with their largest customer and have others in the pipeline:
“We recently won several new projects with our largest customer and we have others in pipeline that are not yet included in our forecast, but which we think are reasonably likely. Several of these new projects are under the second SOW we reported signing with this customer last quarter. We believe that the second SOW potentially gives us access to an even larger generative AI revenue pool with this customer.”
Innodata is a $1.5B market cap stock highly dependent on the announcement of new customer wins and expanding its projects with the largest customer. Our strategy is to watch price closely for the best risk/reward on an entry; and any entry will have a tight stop.
We’ve discussed in the past that Innodata appears to have a low valuation when compared to peers. Our previous analysis also went into the details of Innodata’s products, market positioning and can be found here: “Innodata: Early-Stage AI Data Engineering; Lumpy Growth”
Discovery Members: You are invited to join Knox Ridley on Monday, August 18th at 4:30 pm EST for a special Discovery webinar where he will discuss technical setups on Discovery stocks.a special Discovery webinar where he will discuss technical setups on Discovery stocks.
Revenue up 79.4% YoY
Innodata reported a slight revenue beat in Q2, with revenue coming in 3.6% ahead of estimates to $58.4 million, up 79.4% YoY. This did mark a more than 40 point deceleration in growth sequentially, yet that was expected coming into the quarter.
However, there have been some shifts in forward estimates for Q3 and Q4. For Q3, revenue estimates have been revised nearly $2 million lower since June, now projecting revenue of $59.8 million for growth of just 14.5% YoY. Q4 revenue has been revised more than $5.5 million higher to $70.9 million, for growth of almost 20% YoY. This suggest Q3 is now expected to see be a bottom of sorts before rebounding and reaccelerating into 2026.
For the full-year, Innodata raised its revenue growth guidance from 40% to >45%, now implying revenue of >$247 million, up from $239 million as of Q1. It is important to keep in mind the fluid nature of Innodata’s business, and that any new contractual agreements or expansions could have a large and/or immediate impact on revenue.
Management stated that the guidance hike was driven by “significant new deals that have been finalized since our last call as well as several deals that we believe are highly likely to close in the near term.” Any of these new deals could easily boost growth depending on how quickly they scale.
One key point that could be behind the post-earnings sell-off was some comments made by management around customer engagement and deal sizes. CEO Jack Abuhoff said that with one Big Tech customer, Innodata was “recently awarded a number of significant engagements… enabling us to forecast $10 million of revenue from this customer in the second half of this year,” up from $0.2 million over the last four quarters. Given the size of the revenue guide raise at just $8 million implied, this comment raises some questions about growth in 2H from other core customers.
Largest Customer Accounts for 58% of Revenue
Innodata provided only a brief update on customer progress this quarter, one with its largest customer and the other being the $10 million revenue opportunity discussed above.
With its largest customer, Innodata said that it won “several new projects,” with some of these being under the second scope-of-work (SOW) signed last quarter. Management added that there are other projects in the pipeline with the customer that are not yet included in its forecast, but are “reasonably likely” to be signed in the future.
Innodata’s largest customer accounted for 58% of revenue in Q2 and 59% of revenue for 1H, implying contributions of ~$33.9 million in Q2 and $68.9 million in 1H. This represents no change from Q4 (ie. no expansion) when it was stated the customer was at a $135 million annualized run rate, or ~$34 million quarterly.
Innodata also had no other >10% customers, implying the remaining customers are not spending more than $5.8 million per quarter with the company. The forecast for $10 million in 2H from another customer would likely make said customer Innodata’s second largest and a second >10% customer.
AI Segment Grew 99% YoY on Top of Tough Comps
DDS remains the key driver of Innodata’s growth due to its role in handling AI data preparation, labeling and annotation, AI training and related services.
DDS revenue grew 99% YoY to $50.58 million, though this was slightly down sequentially from $50.83 million in Q1. Although the segment’s growth barely dropped out of the triple-digit range, what’s impressive is that revenue still practically doubled YoY against a 93% growth comp.
Synodex revenue rose 4% YoY to $2.06 million, slowing from nearly 8% growth in Q1.
Agility revenue rose 11.5% YoY to $5.75 million, flat with growth from Q1.
Margins and Adjusted EBITDA expand significantly YoY
Innodata’s margins have significantly expanded YoY, which helps set Innodata apart from stocks in the $1B or $2B market cap range (Innodata is at $1.5B market cap). There was a marginal sequential decline QoQ yet not enough to matter in terms of the overall improvement seen from strong margin and adjusted EBITDA expansion over the past few quarters.
For example, GAAP operating margin was up 14 points YoY from 1.3% to 15.3% in the current quarter.
Q2 GAAP gross margin was 39.4% in Q2, down nearly half a point sequentially but up nearly 11 points YoY. Adjusted gross margin was 42.9%, down slightly from Q1 but up more than 9.5 points YoY.
Q2 GAAP operating margin was 15.3%, up 1.1 points sequentially and up more than 14 points YoY.
Q2 GAAP net margin was 12.4%, down 1 point sequentially but up from approximately 0% in the year ago quarter.
Turning to adjusted EBITDA — Innodata saw a slight sequential improvement in adjusted EBITDA margin in Q2, though on a YoY basis, adjusted EBITDA margin expanded more than 14 points. Innodata had guided for YoY growth in adjusted EBITDA with no further clarity, and has currently reported nearly $26 million for 1H ’25 versus $34.6 million for all of FY24.
Consolidated adjusted EBITDA margin was 22.7%, up from 21.8% in Q1 and 8.6% in the year ago quarter.
DDS adjusted EBITDA margin was 24.2%, up from 22.7% in Q1 and just 5% in the year ago quarter. Tracking DDS’ adjusted EBITDA is important considering the segment accounts for more than 92% of consolidated adjusted EBITDA.
Synodex adjusted EBITDA margin was 22.3%, up from 20.8% in Q1 but down from 26.3% in the year ago quarter
Agility adjusted EBITDA margin was 9.7%, faring the worst out of the three segments as this was down 4 points from Q1 and down nearly 10 points YoY.
EPS Beat by 80% yet EPS expected to decelerate in coming quarters
Innodata handily beat on EPS in Q2, reporting $0.20 in GAAP EPS versus consensus estimates for $0.11. Looking ahead, Q3 EPS estimates are following revenue in moving slightly lower, while Q4 estimates have moved slightly higher. Since our last report, Innodata: Early-Stage AI Data Engineering; Lumpy Growth, here’s how estimates have changed:
Q3 EPS has been revised $0.03 lower, from $0.17 to $0.14, for a YoY decline of more than (73%), though as a reminder the year-ago comp is technically inflated from a $5.9 million income tax benefit.
Q4 EPS has been revised $0.02 higher, from $0.19 to $0.21, for a YoY decline of (31.2%).
For FY25, Innodata is expected to report EPS of $0.76, down (14.6%) YoY before rebounding to 34.5% growth in FY26 to $1.02. To note, FY26’s EPS estimate is still unchanged since June’s writeup.
Cash Flow Declines 60% QoQ
Another blemish in Q2’s report were cash flows, which declined more than (60%) sequentially. Innodata’s balance sheet remains healthy, and the company still has the entirety of its $30 million credit line available.
Operating cash flow was $4.23 million, down (61%) QoQ. OCF margin was 7.3%, down more than 11 points QoQ. For 1H, operating cash flow was $15.1 million, up more than 139% YoY.
Free cash flow was $2.5 million, down more than (70%) QoQ. Free cash flow margin was 4.3%, down more than 10 points QoQ. For 1H, free cash flow was $11.0 million, up nearly 5x YoY.
Cash and equivalents totaled $59.8 million, and debt remained zero.
Deferred revenue was $6.5 million, down from $8.0 million in Q1.
Earnings Call Q&A
There were two topics on the earnings call Q&A session worth highlighting. The first was questions about Scale AI and if Innodata will see more customers onboard as a result of Meta’s large investment of $14.3 billion. The second was why agentic AI will drive more uses cases for the simulation data solutions such as what Innodata and its competitors offer.
Scale AI’s investment is a potential/speculative tailwind:
We covered why Scale AI’s investment could be a potential/speculative tailwind in our prior analysis stating: " Following Meta’s investment, it was rumored that Google, OpenAI and Tesla are looking elsewhere to avoid strengthening Meta at the cost of their proprietary data. Although it’s speculative, the exodus of major players from Scale AI could become a tailwind for Innodata. “
An analyst asked something similar on the Q2 earnings call. The CEO remained vague yet did state that something could materialize in the next couple of months:
“George Frederick Sutton, Craig-Hallum Capital Group:
Nice results. Congratulations. So I wondered if we could talk about during the quarter, your largest competitor, Scale AI was a large majority purchased by Meta. And we've had a few of the large tech companies come out and say they would no longer work with Scale AI. These ostensibely would be tech companies that you have statements of work with. So I'm just curious if you can kind of give us the after effect of that acquisition as you've seen it.
Jack S. Abuhoff, President, CEO & Director
[…] We have, in light of this stepped up that effort with certain companies and there are certain conversations that are going on and are now planned to be happening over the next couple of months that I think could be very exciting for us. I don't know that I can get into particulars much beyond that, but I'll reiterate that we do see an opportunity to accelerate our market presence.”
Agentic AI to drive forward major use cases:
Although agentic AI may still be a few years out before it’s commercially viable, the CEO pointed out the future of large language model (LLM) improvements lies in the quality of data. In order to have agentic AI that can tackle multivariant problems, training data will need to be supplemented with simulation training data.
The CEO stated the following on the call:
“We believe agent-based AI is going to serve as the cornerstone technology that unlocks the full value of large language models and generative AI for enterprises. Moreover, we believe that progress on Agentic AI is likely to soon result in a ChatGPT moment for robotics. Within the next several years, we believe Agentic AI will be served at the edge in hardware devices with which we will commonly interact in many respects in our lives. We believe the market for simulation data services and evaluation services to drive Agentic AI and robotics is likely to dwarf the market for frontier model post-training data.”
Conclusion:
We pointed out in our original analysis that Innodata has many competitors, and we provided an overview of how Innodata must find a path to compete with AI-native startups for data-as-a-service for AI training data. As proven by Scale AI, the market for data labeling and supervised learning is only going to grow in importance – but will Innodata be able to compete? That is a question this earnings report did not answer for investors. For an opportunity like this, we rely heavily on technicals.
Discovery Members: You are invited to join Knox Ridley on Monday, August 18th at 4:30 pm EST for a special Discovery webinar where he will discuss technical setups on Discovery stocks.a special Discovery webinar where he will discuss technical setups on Discovery stocks.
Damien Robbins, Equity Analyst at I/O Fund contributed to this analysis.
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.
