Broadcom is firing on all cylinders and this earnings report cemented the company as number two in terms of AI revenue. It’s not only the AI revenue that sets Broadcom apart, but also its developing software strategy with VMWare.
The headline numbers don’t help to translate underlying AI strength as Broadcom reiterated its full year guidance yet raised AI revenue. This is because some of Broadcom’s segments are coming in lower than expected, while AI is coming in higher than previously guided.
This comment kicked off the tone of the call: “I know we told you in December, our revenue from AI would be 25% of our full year semiconductor revenue. We now expect revenue from AI to be much stronger, representing some 35% of semiconductor revenue at over $10 billion.” This is up from $7.5 billion expected this year, and also up from a $6 billion run rate last quarter ($1.5B per quarter). The AI revenue is roughly 70% ASICs and 30% AI Networking.
In addition to stronger-than-expected AI revenue, Broadcom is expecting dramatic, sequential growth in software bookings, which are expected to grow about 70% QoQ. We need another quarter or two to verify if the rapid growth from VMWare Cloud Foundation will continue, but management implies it will continue to be strong. If so, Broadcom is quickly asserting itself as a leader in AI software as consolidated bookings are expected to add $1.2 billion QoQ from $1.8 billion this quarter to $3 billion next quarter.
Financials Overview:
Revenue and EPS:
Q1 revenue was $11.96 billion, beating estimates by $240 million, and representing YoY growth of 34%. Excluding VMWare, revenue growth was 11% for a 7 percentage point acceleration over the past two quarters, at 4% in the October quarter and 4.9% growth in the July quarter.
Q4 adjusted EPS was $10.99, beating estimates by $0.57. GAAP EPS was $2.84, compared to $8.80 in the year ago quarter.
Broadcom reiterated its fiscal year revenue guide of $50 billion and full year EBITDA guidance of 60%. This compares to an EBITDA margin of 63% to 65% in previous quarters.
Margins:
Q1 GAAP gross margin was 61.7%, compared to 67.4% in the year ago quarter. Amortization of acquisition-related intangible assets adversely impacted gross margin by ~1150bp in the quarter. Adjusted gross margin was 75.4%, compared to 73.8% in the year ago quarter.
Q1 GAAP operating margin was 17.4%, compared to 46% in the year ago quarter. The operating margin was mainly lower due to the increase of amortization of acquisition-related intangible assets, restructuring charges, and stock-based compensation. Adjusted operating margin was 57.1%, compared to 60.9% in the year ago quarter.
Q1 GAAP net margin was 11.1%, compared to 42.3% in the year ago quarter. The net margin was mainly lower due to the increase of amortization of acquisition-related intangible assets, restructuring charges, and stock-based compensation. Adjusted net margin was 43.9%, compared to 50.3% in the year ago quarter.
Cash and Debt:
Q1 operating cash flow was $4.82 billion, representing a 40.3% margin.
Q1 free cash flow was $4.69 billion, representing a 39.2% margin. Excluding restructuring and integration spend of $658 million, free cash flow was 45% of revenue.
Cash, equivalents and short-term investments totaled $11.9 billion.
Debt totaled $75.9 billion. The debt increased from the $39.2 billion in the previous quarter due to the additional debt taken to finance the VMware purchase and the company also assumed $8.3 billion VMware’s debt. We had discussed this in our deep-dive here. The average coupon-rate and years to maturity of fixed rate debt of $48 billion is 3.5% and 8.4 years, respectively. The average coupon-rate and years to maturity of floating rate debt of $30 billion is 6.6% and 3 years, respectively. This week, the company repaid $2 billion of floating rate debt and intends to maintain this quarterly repayment throughout FY2024.
In Q1, Broadcom paid stockholders $2.4 billion of cash dividends based on a quarterly common dividend of $5.25 per share. The company repurchased $7.2 billion of common stock and eliminated $1.1 billion of common stock for taxes due on vesting of employee equity, resulting in the repurchase and elimination of approximately 7.7 million AVGO shares. The Q2 non-GAAP diluted share count is expected to increase to approximately 492 million as the shares issued including VMWare.
Days sales outstanding were 41 days in the first quarter compared to 31 days in the fourth quarter on higher accounts receivable due to the VMware acquisition. This is due to the accounts receivable from VMware having payment terms of 60 days compared to Broadcom’s 30 days.
The company ended the first quarter with inventory of $1.9 billion, up 1% sequentially.
Key Segments:
Software Revenue:
Management reiterated their software revenue guidance of $20 billion this year.
Q1 Software segment revenue of $4.6 billion was up 156% year-on-year and included $2.1 billion in revenue contribution from VMware. In the previous quarter, software was $1.97 billion. This implies 27% QoQ growth in software after stripping out VMWare. When asked about this, management said to not get too excited about this particular growth as it’s due to strong contract renewals. Instead, the CEO explicitly stated: “Yes, don't get too excited over that. So that has also accelerated, but that's not the star of this show, Stacy. Star this show is the accelerating bookings and backlog we are accumulating on VMware.” In fact,it was indicated that some of this could fall off in future quarters given the software guide was not raised.
What the CEO is referring to as the star of the show is the consolidated bookings in software, which grew sequentially from less than $600 million to $1.8 billion in Q1 and is expected to grow to over $3 billion in Q2. Per management: “Revenue from VMware will grow double-digit percentage. Sequentially, quarter-over-quarter, through the rest of the fiscal year.”
Management stated the rapid growth from the VMWare segment is because: “We are focused on upselling customers, particularly those who are already running their compute workloads with vSphere virtualization tools to upgrade to VMware Cloud Foundation, otherwise branded as VCF […] VMware and NVIDIA entered into a partnership called VMware Private AI Foundation, which enables VCF to run GPUs. This allows customers to deploy their AI models on-prem. And wherever they do business without having to compromise on privacy and data — in control of their data. And we are seeing this capability drive strong demand for VCF, from enterprises seeking to run their growing AI workloads on-prem.”
We covered the VMWare acquisition recently in our Broadcom deep dive here.Broadcom deep dive here.
Semiconductor Revenue:
Semiconductor solution sales increased 4% YoY to $7.39 billion, a slight uptick from 3.3% growth in the prior quarter. Stronger-than-expected growth from AI more than offsetting the cyclical weakness in broadband and server storage. According to Bloomberg, this was a bit shy of expectations for $7.7 billion in revenue. This was most likely due to weak wireless, broadband, and server storage segments.
Q1 networking revenue of $3.3 billion grew 46% year-on-year, representing 45% of semiconductor revenue. Management stated the following: “For fiscal 2024, given continued strength of AI NAND working demand, we now expect networking revenue to grow over 35% year-on-year compared to our prior guidance for 30% annual growth.”
Q1 wireless revenue of $2 billion decreased 1% sequentially and declined 4% year-on-year representing 27% of semiconductor revenue. Wireless is expected to be flat YoY for FY2024.
Q1 server storage connectivity revenue was $887 million or 12% of semiconductor revenue, down 29% year-on-year. The company revised its server storage revenue to decline in the mid-20 percentage range compared to prior guidance for a decline in the high teens.
Broadband Q1 revenue declined 23% year-on-year to $940 million and represented 13% of semiconductor revenue. Broadcom revised its outlook for fiscal '24 broadband revenue to be down 30% year-on-year from prior guidance of down mid-teens year-on-year.
Q1 industrial resales of $215 million declined 6% year-on-year. Management stated that industrial resales will be down high single digits this year.
Earnings Call:
Right out the gate, an analyst asked about the surprising acceleration in QoQ bookings on software. Because it’s QoQ, this isn’t accretive software growth from the acquisition (that’s the $600M to $1.8B), rather next quarter represents new, accelerated growth from $1.8B to $3B. The comment from management that “revenue from VMware will grow double-digit percentage. Sequentially, quarter-over-quarter, through the rest of the fiscal year’ helped to solidify that we are already seeing VMWare’s contribution accelerate. We need a few more quarters to figure out if this is a pull forward of some kind and to see where the growth rate will eventually settle. Certainly, it’s off to a promising start.
Question Harsh Kumar (Analyst)
Once again, tremendous results and tremendous activity that you guys are benefiting from in AI. But my question was on software. I think if I heard you correctly, Hock, you mentioned that your software bookings will rise quite dramatically to $3 billion in 2Q. I was hoping that you could explain to us why it would rise almost 100% up, if my math is correct, in 2Q over 1Q. Is it something simple? Or is it something that you guys are doing from a strategy angle that's making this happen?
Answer Hock Tan (Executive)
As I indicated, with the acquisition of VMware — we're very focused on selling, upselling and helping customers, not just buy but deploy this private cloud what we call virtual private cloud solution or platform on their on-prem data centers. It has been very successful so far. And I agree it's early innings still at this point. We just have closed on the deal — well, we closed on the deal late November, and we are now March, early March.
So we had the benefit of at least 3 months, but we have been very prepared to launch and focus on this push initiative on private cloud, VCF. And the results has been very much what we expect it to be, which is very, very successful.
This was also stated later in the call by the CEO:
“All that focus is on the largest, I would say, 2,000 strategic customers. These are guys who want to still have significant distributed data center on-prem […] today's environment, most of these customers do not have an on-prem data center that resembles what's in the cloud, which is very high availability, very low latency, highly resilient, which is one we are offering with VMware Cloud Foundation of VCF. It's exactly replicate what they get in a public cloud. And they love it. Now 3 months. But we are seeing it in the level of bookings we are generating over the last 3 months.”
Near the Semiconductor Trough
Regarding Broadcom’s underperforming segments, the company reiterated what was heard in the Marvell call that we are near the bottom.
Karl Ackerman (Analyst)
Hock, weakness in broadband, server and storage customers is understandable given what your peers have said this earnings season. But perhaps you could speak to the backlog visibility you have with your customers in those markets that would indicate those markets could begin to order again and see sequential growth in the second half through our calendar year?
Answer Hock Tan (Executive)
You're correct. We are — as I say, we are almost like near the trough. This year, '24, first half, for sure, will be the trough. Second half 24, don't know yet. But I tell you what, we have 52-week lead time, as you know. We are very disciplined in sticking to it. And based on that, we are seeing bookings lately, significantly up from bookings a year ago.
Conclusion:
Broadcom’s PS valuation is quite high at 16 compared to the 3-year median of 8. The PE ratio of 42 compares to a 3-year median of 28.5. How to approach this position is not easy given the report was quite strong and there is reason to believe Broadcom will end this year with more beats/raises now that VMWare Cloud Foundation is rapidly accelerating and we already got a $2.5B raise in AI revenue. It’s likely we close Marvell tomorrow and re-allocate some to Broadcom, while taking our time to find an additional entry in the coming weeks.
This article was originally published on Forbes on Feb 29, 2024, 09:34pm ESTForbes Forbes on Feb 29, 2024, 09:34pm EST
The Magnificent 7, defined as Apple, Alphabet, Amazon, Meta, Microsoft, Nvidia and Tesla, have seen a “magnificent” run fueled by AI optimism over the past fourteen months. The Magnificent 7 returned more than 106% in 2023, doubling the Nasdaq 100’s nearly 54% gain and significantly outperforming the S&P 500’s 24% gain. At first glance, it may appear that the Magnificent 7 are continuing their outperformance of the broader indexes in 2024.
However, like dominos falling, these market generals are topping out and diverging from the broad market. First Tesla in July of 2023, then Apple and Google in February have topped, and now Microsoft is not making a new high with the broad markets’ most recent run higher.
The Magnificent 7 of 2023 have now become 2024’s Magnificent 3: Nvidia, Meta and Amazon. Of these, Nvidia’s saw a stellar start to the year as shares have gained nearly 60% YTD due to the GPU leader’s beat-and-raise quarters.
Source: TradingView
There are two reasons why this matters – which we also outlined in our analysis “Five Stocks (Not Seven) Can Lead to New Highs” from October – that “a handful of these stocks [the Mag 7] can push the bigger markets higher,” but now we’ll need more than just three to keep the rally going.
First, these 7 stocks hold a significant weighting within the indexes. It will be difficult for a sustained push higher to continue if these FAANGs do not participate, considering their outsized weighting.
The Mag 7 comprises more than 40% of the Nasdaq 100 and more than 29% of the S&P 500.
MSFT, GOOGL, AAPL, and TSLA account for about 18% of the S&P 500 and about 25% of the NASDAQ-100.
For reference, just Apple and Microsoft combined hold a larger weighting in the S&P 500 than Berkshire Hathaway, JP Morgan, UnitedHealth Group, Visa, Exxon, Mastercard, Johnson & Johnson, Procter and Gamble, Home Depot, Costco, Merck, and Chevron combined. If these companies collectively all stalled, it would be a major warning sign. Yet, Apple and Microsoft are both stalling.
Sign up for I/O Fund's free newsletter with gains of up to 221% – Click hereClick hereClick here
Secondly, when the cycle leaders start to underperform, it tends to mark the start of a trend change. The FAANGs have been the undoubted leaders of this bull run, and we are now seeing them start to trend lower against the indexes. More times than not, the leaders on the way up, tend to be the leaders on the way down.
In today’s bull cycle, this leaves Nvidia, Meta and Amazon as the three remaining generals making new highs with the markets.
Nvidia, Meta and Amazon are the three remaining generals making new highs with the markets. Source: TRADINGVIEW
Combined, the trio account for approximately 15.8% of the Nasdaq 100 and 10.8% of the S&P 500. Nvidia’s post-earnings surge, in which the chip giant added nearly $250B in value, helped the S&P 500 add more than $2 trillion in market cap as it boosted other AI and tech stocks in general. Should the trio begin to follow in the path of the four fallen dominos, setting a high and drifting lower, the market may be at risk of giving up some of its newfound gains, similar to what we had discussed in our analysis “Apple Can’t Save This Tech Rally” at the end of January. In this, we outlined how both the bull and bear cases for the market “are calling for a level of volatility in 2024 that will, at least, retrace the rally we’ve seen since November 2023.”
Concentration Risk Elevated
To an extent, the narrow leadership of this market stemming from the Magnificent 7’s AI-powered gains has raised warning bells for some investors, as the market’s concentration has surpassed levels seen in the dot-com bubble. To be clear, my firm is a pioneer in building an AI portfolio, and a selloff would be a buying opportunity. However, narrow leadership is a problem not to be ignored, and this is best illustrated by the chart below:
Source: CME
As mentioned earlier, the Magnificent 7 account for more than 29% of the S&P 500, more than the 21% concentration of the top 7 stocks in the S&P 500 seen in 1999 and 2000 — keep in mind that Tesla is no longer one of the top 10 largest stocks in the S&P 500, so the concentration of the top 7 today is above 30%. This also marks a dramatic increase from the 14% concentration seen a decade ago.
What this means is that as the Magnificent 7 as a whole continue to outperform – the seven have already gained more than 22% YTD in 2024 – they will continue to cover up the turbulence in the broader market that is brewing under the surface. For example, at the end of February, the Nasdaq 100 and S&P 500 are up nearly 9% and over 7%, respectively, while the equal-weighted S&P 500 has gained just over 2%.
Source: TradingView
This concentrated dominance has helped the S&P 500 push to new highs, more than 6% above its 2021 high, while the equal weight S&P (orange) has yet to reclaim that 2021 high, sitting about 100 points lower. The influence of the Magnificent 7 is clearly visible — the S&P 500 has a 26 percentage point outperformance of the equal-weight index, returning 81% versus 55% over the past five years; this gap has widened throughout 2023, from 8 percentage points in April to 14 percentage points in July to 20 percentage points in October.
Source: YCharts
I/O Fund Portfolio Manager Knox Ridley outlined in our analysis in October, 5 Stocks (Not 7) Can Lead To New Highs that “a handful of these stocks [the Mag 7] can push the bigger markets higher, and even potentially make another high in the NASDAQ-100.” The setup was that the indices were “due for a sizable bounce over the coming weeks – months, which we believe will be led by a handful of Big Tech names.” Now that we are at new highs, we think we will need more than just three of the Mag 7 to keep going.
Every Thursday at 4:30 pm Eastern, the I/O Fund team holds a webinar for premium members to discuss how to navigate the broad market, as well as various stock entries and exits. We offer trade alerts plus an automated hedging signal. The I/O Fund team is one of the only audited portfolios available to individual investors. Learn more here.Learn more hereLearn more here.
Valuations Relatively Intact
Though the recent momentum-filled surges in AI favorites including Super Micro and Nvidia have some investors drawing parallels to Cisco’s ascent in 2000, valuations for the Magnificent 7 are relatively intact.
Tesla is struggling with earnings growth as price cuts bite margins, while Apple’s growth headwinds are leading to minimal earnings growth; on the other hand, Amazon is showing strong earnings leverage from improvements in its margins, Google is trading at a near 30% discount to its year-ago PE of 30x, and Nvidia is eerily cheaper now than it was when it had bottomed in October 2022 in the low $100 range.
Source: YCharts
Compare this to Cisco, given the parallels being drawn, which traded at more than 150 times earnings at the peak of the dot-com bubble – or more than twice as high a multiple as the most expensive of the Mag 7 of today.
We discussed on Fox Business News this week that keeping an eye on valuation is important for determining which stocks to buy on dips. The impact AI has had is very visible on the top line with blowout quarters from Nvidia, and on the bottom line with blowout quarters from both Nvidia and Meta. However, AI’s impact on valuations is being overlooked as these valuations are low and setting up a new buying opportunity should the broad market present weakness.
Conclusion
We will continue to track how the Magnificent 3 perform over the next few weeks, and whether Meta, Nvidia, and Amazon will continue to lead or if they will follow the trend of the remaining four in underperforming versus the broader indices.
When these cycle leaders start underperforming, it usually marks the start of a trend change. The FAANGs undoubtedly have led this bull run since 2023. We are now looking for what will lead the market next, and most importantly, when.
If you own AI stocks or are looking to own AI stocks, consider joining us for our next broad market webinar. Every Thursday at 4:30 pm Eastern, the I/O Fund team holds a webinar for premium members to discuss how to navigate the broad market, manage risk, as well as revealing our various long-term game plans regarding stock entries and exits. We offer trade alerts plus an automated hedging signal. The I/O Fund team is one of the only audited portfolios available to individual investors. Learn more here.
I/O Fund Portfolio Manager Knox Ridley and I/O Fund Equity Analyst Damien Robbins contributed to this report.
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.
Broadcom has greatly outperformed the FAANGs over the past decade yet is rarely discussed as one of the market’s top-performing stocks. The ticker certainly does not participate in a catchy acronym. Half the battle with Broadcom is there are many revenue segments to analyze, some go through harsh cyclical downturns, and it acquires companies hand over fist. To put it plainly, this is not an easy stock to cover; there are no pithy ways to summarize the products and it doesn’t offer growth stock qualities.
Broadcom is reporting the second highest AI revenue on the stock market today, accounting for 16.1% at $1.5 billion, up from the July quarter at $1 billion and 11.3% of revenue. Of the smaller players, peers like AMD have guided for AI revenue of $3.5 billion in 2024, which would account for about 13.6% of the 2024 estimated revenue. Marvel expects AI revenue of $400 million in 2023 or 7% of revenue and is rising to $800 million or 13% of the expected 2024 revenue. Therefore, Broadcom is in second place, but notably, does not have a large lead by percentage of revenue. Juniper at 23.5% of revenue has a higher percentage yet is being acquired by HPE.
Ultimately, our goal is to get this stock lower, but to put into motion now the in-depth research required for a potential entry. As stated in the 2023 Year in Review webinar, Meta was the one that got away given its bottom line fits our criteria, but I consider Broadcom the one sitting in plain sight.
Broadcom has many departments that have been strung together through acquisitions with a vision of consolidating Ethernet, ASICs and now virtualization under one company. The reason this company is sitting in plain sight is because it was a major winner from the mobile era, is partnered with Big Tech as we move in the AI era, and is putting together the pieces to participate in AI strategically by not needing to compete on GPUs.
Broadcom’s Switch Products and Switch Fabric
Broadcom has a dominant market share of switching and routing semiconductors for hyperscalers and is seeking to maintain its market share, most especially as AI changes networking requirements. The Jericho3-AI launch was last April, which is a redesign intended to compete with Nvidia’s InfiniBand.
Broadcom has three switch products. The Tomahawk is the high-bandwidth switch platform, Trident is the platform with more features, and the new Jericho line combines the Jericho switch with routing ASICs. The Jerico3 was redesigned with deep packet buffers. Tomahawk and Trident are used in data centers yet are not optimized for AI workloads, especially when compared to InfiniBand.
Jericho has 160 switch ports dedicated to switch fabric, which allows multiple ASICs to be stitched together to support GPU clusters. The asymmetric split helps the chip overcome network congestion and network failures. According to Broadcom, Jericho3-AI performed 10% better than “alternative network solutions” — which is a clear reference to InfiniBand.
A few specs before we go deeper into how Broadcom’s solutions compare to Nvidia’s. As you’ll note, the specs for InfiniBand are superior with support for 64X 400Gbp or 128 200 GB/s ports compared to Jericho’s 36X 400 GbE and 72X 200GbE network-facing ports.
Jericho has 144X SerDes lanes and 106Gpbs PAM4 supporting 18X 800Gbe, 36X 400 GbE and 72X 200GbE network-facing ports. The Jericho3-AI allows for more than 32,000 GPUs to be linked for a massive AI training system and links directly to GPUs without the need for a server bus.
Tomahawk5 runs at 100 GB/second with PAM4 with aggregate bandwidth of 51.2 Tb/s
Compare this to Nvidia’s Quantum-2 InfiniBand which has support for 64X 400Gbp or 128 200 GB/s ports with 51.2 Tb/s and 66.5 billion packets per second.
Nvidia’s new Spectrum X Platform is an Ethernet solution that delivers 1.6X better networking performance than traditional Ethernet with 256X 200 Gbe ports or 16,000 ports for larger training systems.
On a similar note, and while we are on the topic, Marvell has some skin in the game too by offering a 51.2T switch called Teralynx 10 that offers ultra-low-latency at 80% cost savings. According to Moor Insight Strategy, Marvell is the supplier for AWS for “electro-optics, networking, security, storage, and custom-designed solutions.”
Marvell’s Nova 1.6 Tbps PAM4 electro-optics have eight 200G lanes that “double the networking bandwidth while reducing power and cost per bit by 30%.” According to the press release, “by doubling the bandwidth per lambda, the Nova-based modules reduce the number of lasers and related optical components by 50%.”
Marvell hopes to help data centers transition to 51.2 Tbps networking architectures by offering a platform that needs 32 optical modules instead of 64 optical modules.
Here’s a a quick glance on the rankings for AI networking revenue (approx. revenue)
Nvidia in first place with $2.5B per quarter from InfiniBand
Broadcom in second place with $1.5B from AI, primarily networking
Marvell at $200M per quarter from AI, primarily networking
Juniper Networks reported $321.2 million in the AI enterprise segment, or 23.5% of revenue. Recently, it was announced that Juniper is being acquired by HPE.
Further out in FY2025 and/or CY2025:
Cisco is expecting $1 billion in orders from a recent Nvidia partnership, per the earnings call: “our expectation is the majority of that $1 billion in orders will turn into revenue in our fiscal '25, just to be clear.”
Arista is expecting $750 million by 2025: “We are cautiously optimistic about achieving our AI revenue goal of at least $750 million in AI networking in 2025.”
AI Networking
We’ve covered networking as it relates to data centers, 5G, cloud applications and enterprises when we wrote about Nvidia’s acquisition of Mellanox in 2020 and Marvell’s acquisition of Inphi.
A few years back, we discussed that Nvidia acquired Mellanox for the strategic synergy that InfiniBand and Ethernet can provide in boosting GPU performance. Without proper interconnection, GPU performance could be limited, and so Nvidia strategically wanted to create the best-case scenario of owning both markets for AI accelerators — and their fabric and interconnects.
Mellanox supports Virtual Protocol Interconnect (VPI), which allows the ubiquitous Ethernet to provide bandwidth as cheap as possible, and InfiniBand to deliver higher throughput and fewer bottlenecks during high loads. In 2019, the split in Mellanox’s revenue was about 40% InfiniBand and 60% Ethernet. By leveraging a hybrid of Ethernet and InfiniBand, Mellanox was able to take market share from Ethernet incumbents.
The acquisition went under review in China, with officials believing Mellanox’s market share at the time was about 55% to 60% of the global interconnect market and 80% to 85% of the Chinese market. This illustrates how popular Mellanox was before Nvidia acquired the company.
The outcome of the review was that Nvidia was required to decouple the sale of InfiniBand from the sale of its GPUs to where a customer could buy one but not the other at no penalty. Even if a customer can buy them separately, there are many cases where it’s not practical to do so, such as with the DGX and HGX systems which achieve optimal performance with the A100s/H100s and InfiniBand.
Nvidia has stated that InfiniBand increase the effectiveness of AI infrastructure by 20% to 30%. Remote Direct Memory Access (RDMA) reduces CPU overhead by offloading data movement to network adaptors. In addition, Ethernet has quality of service (QoS) flow control and advanced error handling mechanisms that increase its network efficiency capabilities.
In the last earnings report, Nvidia stated in the opening remarks that “Networking now exceeds a $10 billion annualized revenue run-rate. Strong growth was driven by exceptional demand for InfiniBand, which grew fivefold year-on-year […] Azure uses over 29,000 miles of InfiniBand tabling, enough to circle the globe.”
The software defined fabric is popular for its low latency as its architecture reduces packet loss, high bandwidth and low management costs. The high-speed data transfer has link speeds of 10 to 400 gigabits per second due to its low overhead and efficient transport protocols, which is why InfiniBand is adopted by supercomputers and also AI/Big Data applications with high performance clusters. Due to very low latency, InfiniBand delivers real-time data transfer.
The $10 billion in annualized revenue run-rate reported by Nvidia in the Q3 October quarter represented 500% growth, which is nearly double the growth rate of the overall data center. Per the Q3 earnings call: “Networking now exceeds a $10 billion annualized revenue run-rate. Strong growth was driven by exceptional demand for InfiniBand, which grew fivefold year-on-year [..]” At the time, the data center had a run rate of $60 billion, so networking was 16.6%. In Nvidia’s most recent quarter, networking grew 155%.
According to Del’Oro, a research company out of the UK, AI systems account for less than 10 percent of the total addressable market for network switching, and of that, 90 percent are using Nvidia/Mellanox InifiniBand due to InfiniBand reducing packet loss, which is ideal for AI training workloads.
If right now, you’re thinking: “I thought this analysis was about Broadcom, not Nvidia!?” then that’s a fair question. Given AI networking is heating up across the board (Nvidia’s run rate, Broadcom, Juniper/HPE, Marvell, Cisco potentially, Arista), it’s important we touch on why Infiniband owns 90% of the AI market right now. We are overdue on revisiting Mellanox/InfiniBand as it’s been four years since we covered the acquisition. This also helps frame how Broadcom intends to compete with Ethernet.
Ethernet Vs InfiniBand
I wish I could make networking more conversational, but it’s pretty challenging to do that! Here are some bullet points on how the two compare; I’ve bolded the more important takeaways:
Benefits of Ethernet:
Raw bandwidth is a benefit with Ethernet hitting 51.2 Tb/s two years ago with support for 800 Gb/s port speeds. InfiniBand lags by topping out at 51.2 TB/s with 400 Gb/s port speeds. Although typical server nodes do no need the extra bandwidth, AI clusters come with 400 Gb/s NIC per GPU with some nodes having four to eight GPUs. By 2025, Dell’Oro believes switch ports for AI networks will be operating at 800 Gb/s and will further double to 1600 GB/s by 2027. Dell’Oro believes switch ports for AI networks will be operating at 800 Gb/s and will further double to 1600 GB/s by 2027.
smartNICs and AI-optimized switch ASICs help to reduce packet loss
Large pool of vendors whereas InfiniBand increases dependency on Nvidia.For this reason, AWS and Google Cloud have remained on Ethernet as they prioritize custom silicon.
Ethernet is the incumbent networking standard and most cloud providers have invested heavily here already. With Ethernet, providers don’t have to manage a new network stack.
Ethernet switching has evolved to where a new term has been coined “lossless” Ethernet. Even Nvidia is moving in this direction with their Spectrum X platform, due out this year.
Benefits of InfiniBand:
Outperforms for AI/ML workloads due to low latency and by reducing packet loss. Data packets are sent in a serial approach so multiple channels of data can be sent simultaneously. This is much better for AI/ML than a parallel approach for internal data flow, which creates bottlenecks.
Has 3X to 4X lower latency than traditional Ethernet switches based on ASICs
Highly scalable, can support tens of thousands of nodes per subnet. InfiniBand is also cheaper as it requires fewer connections for reliability.
Its QoS and failover capabilities are a reason it’s adopted for high-performance computing environments.
Reduces CPU resources
So, why is Ethernet Making a Comeback?
Broadcom’s Jericho3-AI has some promising benchmarks that could help shift the dominant market share InfiniBand has in AI networking (or at least prevent a monopoly). These benchmarks showed the Jericho3-AI outperforming InfiniBand by 10%, which is substantial when dealing with AI systems as it’s enough to increase the collective operations of the system.
“Leveraging this unique functionality, the Jericho3-AI fabric provides at least 10 percent shorter job completion times versus alternative networking solutions for key AI benchmarks such as All-to-All. This performance improvement has a multiplicative effect on decreasing the cost of running AI workloads since it implies that expensive AI accelerators are used 10 percent more efficiently. The network, in effect, pays for itself.” — You can read the press release here.
This means bare metal can work more effectively. Per Broadcom: “because it can handle 800Gbps port speed (for PCIe Gen6 servers) and more, it is a better choice [than InfiniBand.” At high price points, all hyperscalers want to see their investments working at maximum clock times. This is achieved by better load balancing and congestion control to improve network latency, whereas InfiniBand reduces port and hop latency inside the switch. Broadcom calls their product differentiation “Perfect Load Balancing” and “Congestion-Free Operation.”
A note on PCIe
The maximum bandwidth supported by PCIe 5.0 is 400Gbps per port. By using 106Gbps PAM4 SerDes, ASICs can be tuned to support 100, 200 and 400 Gbps port speeds. To work around this, and to achieve 800Gbps, chip makers are building NICs directly into the accelerator. According to The Register, the 800Gbps ports built into accelerators may reduce bottlenecks before PCIe 6.0 arrives on the market. The Register, the 800Gbps ports built into accelerators may reduce bottlenecks before PCIe 6.0 arrives on the market. This is what Broadcom is referring to.
Jericho3-AI supports 36 ports at 400 Gbps speed, and this can support Nvidia’s powerful DGX H100, which have 8 ports of 400 Gbps speed. In this case, four node racks are within Jericho’s capabilities. However, the Quantum-2 InfiniBand can handle 64 ports of 400Gbps, and so for Nvidia’s GPUs, it outperforms. Broadcom’s answer to this is that AWS and Google still prefer to not have vendor lock-in with Nvidia and use the Jericho3-AI to make use of their extensive Ethernet systems.
Overview of Custom Silicon (ASICs):
In addition to AI networking, Broadcom participates in the custom silicon market. ASICs are application-specific integrated circuits that are customized to perform a specific function for a specific application, hence the term “application-specific.” This is in contrast to GPUs which are more general-purpose. ASICs are expensive at the onset, yet become cheaper with volume production. We first published this graph in 2019 but the comparison still applies:
For now, custom silicon only makes sense for a company with deep coffers that has immensely popular applications – such as Google, Meta, Amazon. These companies use custom silicon to drive down costs on GPUs for their most popular applications. Across ASICs, the most well-known is Google’s tensor processing unit (TPU).
Google was one of the first to require low-power machine learning workloads for Search, YouTube and Google Maps. The compute intensive workloads were running on Nvidia’s GPUs for both training and inferencing until Google made their own processing unit, TPUs, to perform workloads at a lower cost and higher performance.
Performance between TPUs and GPUs is often debated depending on the current release (A100 versus fourth-generation TPU, for example). In some cases, TPUs have better performance per watt for power-constrained applications. Notably, some of this comes with the territory of being an ASIC, which is designed to do one specific application very well whereas GPUs can be programmed as a more general-purpose accelerator. In this case, the benchmarks where TPUs compete are object detection, image classification, natural language processing and machine translation — all areas where Google’s product portfolio of Search, YouTube, AI assistants, and Google Maps, for example, excels.
Notably, TPUs are used internally at Google to help drive down the costs and capex of its own AI and ML portfolio and they are also available to users of Google’s AI cloud services. For example, eBay adopted TPUs to build a machine learning solution that could recognize millions of product images.
Unless Google releases an internal technology as open-source, it won’t be adopted by the competitors. This is where Nvidia’s neutral position as traditionally a hardware company becomes a positive as it’s universally used by Amazon, Microsoft, Google — — and Alibaba, Baidu, Tencent, IBM and Oracle. Meanwhile, TPUs create vendor lock in (with a direct competitor) which most companies want to avoid. eBay is the exception here as the company needs Google-level object detection and image classification.
Why ASICs are Not a Near-Term Threat to GPUs
AI investors will need to get comfortable hearing about the battle between ASICs and GPUs. This debate has been going on since at least 2018, when Nvidia’s biggest threat was thought to be Google’s custom TPUs. There is some merit to these concerns as the largest customers for Nvidia’s GPUs have enough cash to make custom chips. There’s roughly $35B to $40B per Big Tech company per year that executives will naturally want to optimize to drive down costs.
To program ASICs is difficult, and they are application-specific, which means they cannot be reconfigured. Nvidia is wildly popular because GPUs are easy to program and are the best choice for a wide range of applications. Developers create the moat, which was our original Nvidia thesis. Therefore, I don’t believe there is much risk that Big Tech commercializes AI accelerators.
However, it’s quite plausible that someday Big Tech will reallocate capex toward more ASICs and fewer GPUs to where it will impact Nvidia. For now, demand outstrips supply, and there are long lead times for Nvidia’s GPUs. If a company like Google reallocates to more TPUs, another enterprise will certainly step up to fill those orders.
Broadcom & Google Partnership
It was confirmed last year that Google is a customer of Broadcom for its ASICs (TPUs). This was officially reported when The Information wrote an article stating Google wanted to ditch Broadcom in 2027, which Google has since denied:
“We are productively engaged with Broadcom and multiple other suppliers for the long term. Our work to meet our internal and external cloud needs benefit from our collaboration with Broadcom; they have been an excellent partner, and we see no change in our engagement." -Google’s response to The InformationGoogle’s response to The Information
Prior to this, it was never directly stated that Google was Broadcom’s main ASICs customer. Here is how Broadcom discussed it: “As you know, we supply a major hyperscale customer with custom AI compute engines. We are also supplying several hyperscalers a portfolio of networking technologies as they scale up and scale out their AI clusters within their datacenter.”
The following has also been stated about the Google-Broadcom relationship: “Broadcom supplies wireless chips for Google phones as well as chips for its data center and cloud services. At the same time, Broadcom is one of Google Cloud’s biggest customers for its cloud products. This bidirectional relationship has also forged a special bond." Per the same report, Meta is also working with Broadcom on ASICs, although does not deploy many of these “yet”
Last April, Broadcom migrated its infrastructure from AWS over to Google Cloud. Per the announcement: “Broadcom, a provider of enterprise security solutions, recently worked with Google Cloud Consulting to migrate its infrastructure from Amazon Web Services (AWS), and found the combination of technology and expertise critical for success. “Google's deep technical skills and its data, security and AI offerings have accelerated our transformation towards becoming a software-led company,” said Andy Nallappan, Vice President, CTO and CSO, Broadcom.
VMWare –Software-Defined Networks and Data Centers
In November, Broadcom closed its acquisition of VMware for $69 billion. VMWare is virtualization software that virtualizes compute and data centers. The software creates an abstraction layer, or a “hypervisor” which is the technical term for a computer or server that runs virtual machines called ESX. VMWare was the first company to virtualize x86 machines and was founded in 1998. Operating systems, such as Linux, Windows and MacOS, can share the same, virtualized resources by running on a x86 machine.
Over the past decade, VMware pivoted to offer a software-defined data center. On the most recent earnings call, Broadcom discussed building essentially a private cloud on-premise, which has some advantages, such as lowering capex by pooling memory, security, networking and server resources. “Our strategy going forward is simply to enable global enterprises to run their applications across the other data centers as well as on public clouds by consuming VMware’s higher-value software stack.”
Later it was stated: “We are creating with VMware, the same experience of virtualization of the data center on-prem for those companies, which has workloads, by the way, that are already running VMware products that application that’s already written on VMware Cloud Foundation. This is then giving these enterprises the opportunity to have a hyperscaler on-prem. That’s the plan we’re doing, plain and simple.”
Where software defined networks have seen quite a bit of success is with 5G networks. SDNs separate the control plane on embedded switching systems. This allows the networks to be managed remotely. Products from different suppliers can be used without incompatibility issues. By using open APIs, the 5G market has benefited from a more neutral ecosystem by allowing products from different suppliers. This is because SDNs allow network functions to be programmed by APIs instead of proprietary interfaces.
In 2012, VMWare acquired Nicera to create VMware NSX, virtual networking and security software that virtualizes network components. The NSX products, including NSX-T data center, programmatically creates and manages virtual networks from Layer 2 to Layer 7, which is defined as switching, routing, access control, firewall and QoS. NSX Manager and transport nodes can be assembled in seconds for proof-of-concept deployments, deployments with up to 64 hosts, or large-scale environments. Software-defined networks have a natural synergy with Broadcom, the leader in networking hardware.
A few years ago, VMWare expanded to virtualize containerized workloads for Kubernetes clusters. This product is referred to as Tanzu. This was a necessary evolution to keep cloud native customers. Many Kubernetes clusters use something called a multi-tenancy solution, which is to have non-connected “tenants” use a common pool of resources. This can be hard to implement correctly, and also has limited functionality once it’s set up. Virtualized containers are similar to a single-tenancy solution by having its own API server, controller manager and storage for data. Yet, it’s similar to a multi-tenancy solution by using a common pool of resources. Per the Broadcom earnings call: “And to attract and keep these workloads across the environment, we are investing in a rich catalog of microservices tools. This will be our focus. And the noncore businesses of end-user computing and Carbon Black will be divested.”
All of this sounds good, but virtualization is not a wild success. The software defined data center market at one time was expected to reach $77 billion by 2020 but instead has only reached $28 billion as of 2023. There are many vendors in the space, which creates pricing wars.
And so, it’s speculative as to how the software defined data center or networks would ultimately accelerate in growth based on AI workloads. The anticipated acceleration is mainly from restructuring, rather than product-market fit. This is what management said on the earnings call: “And it just doesn’t stop there because it’s the math and the trajectory. And to answer your question, you’re right, we are accelerating from $12 billion, and we’re probably seeing a double-digit growth for the next three years, just by sheer math of selling that higher value virtualization stack versus the very loose component sales in the past, particularly on compute only.”
There was a solid question about this in the Q&A which I’m quoting in full below.
Harlan Sur:
[…] but given the significant performance requirements of these workloads, right, training, inference, it appears that more of the near-term adoption of running these workloads is on bare metal, GPU, TPU, accelerated servers. So, how is the team exploiting a software-defined data center solutions via either cloud foundations or Tanzu to try to help customers focus on AI sort of drive better utilization, better economics, faster deployments on this very fast growing part of the market?
Hock Tan, CEO:
Well, as you may be aware, in the last VM Explore in Las Vegas, VMware came out and announced in partnership with NVIDIA, the VMware Private AI Cloud Foundation. Another way of describing it is, the VMware Cloud Foundation Software Stack, the whole VCF stack runs NVIDIA coder, runs the NVIDIA GPU. That is the partnership. So, if you’re an enterprise, it’s a very easy step to get into gen AI analytics because the data center that you as an enterprise own on-prem that runs VCF will by default run the NVIDIA GPU software stack as well.
Another way to put it, it virtualizes the NVIDIA GPU. That’s the VMware software stack as well. So it’s a very strong attraction in our — from our perspective to, in fact, accelerate thinking of a lot of enterprise to adopting the whole VCF site. It’s simply because not only does it virtualize the data centers and make your data on-prem data center much more resilient, easier to manage, lower cost to manage, it has the added benefit, a big attraction this is of being able to right away start running AI workloads
Broadcom’s Financial Overview:
Broadcom consistently delivered a net profit margin exceeding 37% throughout fiscal year 2023. Additionally, Broadcom demonstrates exceptional cash flow generation, with free cash flow exceeding 44% in each quarter of FY2023 and even surpassing 50% in the last three quarters.
Broadcom's acquisition of VMware in November 2023 is intended to bolster the company's position in the AI space, while also strengthening its software business. Secondly, the merger will eventually create recurring revenue streams, which once complete, will be a positive. Furthermore, this acquisition diversifies Broadcom's portfolio, with infrastructure software projected to account for roughly 40% of FY2024 revenue compared to 21% in FY2023. This shift mitigates the impact of cyclical downturns inherent to the semiconductor industry.
The integration process is expected to take a year and will initially have a drag on profit margins due to transition costs and VMware's lower margin profile, cost-cutting measures and merger synergies are anticipated to improve margins in the long term.
When an acquisition is complete, it typically weighs on stock price while investors move to the side lines to see how the teams merge internally and also externally for customers. Unfortunately, the VMware acquisition is not going too well with rumors that customers are disgruntled alongside Broadcom spinning off non-core segments and selling them off to other companies.
That complicates the picture as it requires understanding the precise impact of each segment independently, which is impossible to do given cloud companies tend to cross-sell products. It's also important to note that Broadcom is transitioning VMware clients to a subscription-based business. Per the earnings call: “[…] and we are converting more and more customers step-by-step as they come up for renewal into this higher value stack, and we’re doing it on a subscription basis. So become very focused. So we will kick it off at a much lower rate — because subscription generally brings down revenues, as you know, in software based on revenue recognition. But we see a trajectory of accelerated growth even in 2024 — through 2024. And it just doesn’t stop there because it’s the math and the trajectory. And to answer your question, you’re right, we are accelerating from $12 billion, and we’re probably seeing a double-digit growth for the next three years, just by sheer math of selling that higher value virtualization stack versus the very loose component sales in the past, particularly on compute only.”
When looking at revenue growth, it’s important to strip out VMware’s contribution post-acquisition. Management is firm in the earnings calls that the VMware will accelerate. If this comes to fruition, the Street will likely reward Broadcom as VMware is the primary risk given the synergy of the rather large acquisition ($60 billion) is unproven. However, due to the uncertainty around restructuring the VMware acquisition, time is on our side to try to get AVGO at a lower price.
Broadcom’s Revenue and EPS:
Broadcom’s revenue in the Q4 FY2023 ending Oct grew by 4.1% YoY to $9.3 billion. Revenue was in line with the analyst consensus.
Analysts expect revenue to grow 31.6% YoY to $11.73 billion in the next quarter. Since the company completed the acquisition of VMware on November 22, 2023, these estimates include VMware’s revenue. Headline revenue numbers are expected to accelerate for the next four quarters due to VMware and then will re-acclimate in the January 2025 quarter at 16.4% growth. These quarters are likely to be watched closely due to reasons outlined above, which is that it will be apparent and quite easy to model VMware’s impact.
Organic revenue, excluding the VMware acquisition, represents a 6.1% YoY growth for FY2024 ending in October, down from 7.9% in FY2023 due to the cyclical slowdown in the semiconductor sector. See more discussion on this below.
However, the company will see a higher growth rate in FY2025 at 10.9% YoY for $55.28 billion. The growth rate is helped by an increasing contribution from AI. Management stated in the earnings call that revenue from generative AI will grow from 15% in FY2023 to more than 25% in FY2024. The company’s CEO, Hock Tan, said in the earnings call, “Revenue from generative AI in fiscal ‘23 reached 15% of semiconductor revenue, in line with our expectation. And moving on to fiscal ‘24, we forecast semiconductor solutions revenue to be up mid- to high-single-digit percent year-on-year. We expect revenue from generative AI to represent more than 25% of the semiconductor revenue, consistent with prior guidance, which more than offset the lack of growth from non-AI semiconductor revenue.”
Management chose to not provide quarterly guidance, and to instead provide FY2024 guidance of $50 billion, representing YoY growth of 39.6% at the mid-point. This is a break in style as quarterly guidance is typically given. Here is what was said on the call: “Now on to guidance. As Hock discussed, with the recent closing of our VMware acquisition and the integration process, which will take at least one year, for fiscal 2024, we will provide our outlook for the full year instead of quarterly guidance. Based on current business trends and conditions, our guidance for fiscal year 2024 is for consolidated revenues of $50 billion. Within this, our fiscal year 2024 semiconductor revenue is expected to grow mid- to high-single-digit percent year-on-year. Our fiscal year 2024 infrastructure software segment revenue from continuing operations is expected to be $20 billion, including $8 billion from CA, Symantec Enterprise and Brocade and $12 billion from VMware.”
Management mentioned that VMware’s 11 months expected contribution from the time the acquisition is closed is $12 billion for the FY2024 ending October.
Pictured Above: Revenue includes VMware acquisition. Organic revenue, excluding the VMware acquisition, is expected to be 6.1% YoY growth for FY2024 ending in October, down from 7.9% in FY2023.
Segments
The Semiconductor Solutions segment revenue grew by 3% YoY to $7.3 billion and has witnessed a cyclical slowdown. It is down from 5% growth in the previous quarter and 26% growth in the same quarter last year.
In this segment, networking revenue is the largest by end markets, constituting 42% of Q4 semiconductor revenue.
Networking revenue grew by 23% YoY to $3.1 billion due to strong demand from hyperscalers. Due to AI, management expects FY2024 networking revenue to grow 30% YoY, up from 21% in FY2023.
The company is a beneficiary of generative AI and reported $1.5 billion in Q4 FY2023, for a run rate of $6 billion per year. Citi Analyst, Christopher Danely, said in a research note that the company’s AI revenue will double from $4 billion in FY2023 to $8 billion in FY2024, and he expects the AI business will offset the correction in the semi-business.
Looking further out, Mizuho analyst Vijay Rakesh said “that Broadcom’s AI revenue will likely grow from $8 billion in 2024 to $20 billion in the calendar year 2027, thanks to its custom ASIC AI portfolio.”
The company’s CEO, Hock Tan, said in the earnings call, “This was primarily driven by strong demand from hyperscalers for our custom AI accelerators and as well for our networking switches, routers and NICs, Network Interface Cards, dedicated towards scaling our AI data centers.”
“As you know, even as Ethernet is the standard protocol in front-end networks, hyperscalers are also deploying Ethernet predominantly in their AI networks. In fiscal ‘23, networking revenue grew 21% year-on-year to $10.8 billion. If we exclude the AI accelerators, networking connectivity represented about $8 billion, and this is purely silicon, not systems, not cable nor subsystems. In fiscal 2024, we expect networking revenue to grow 30% year-on-year, driven by accelerating deployment of networking connectivity and expansion of AI accelerators in hyperscalers.”hyperscalers are also deploying Ethernet predominantly in their AI networks. In fiscal ‘23, networking revenue grew 21% year-on-year to $10.8 billion. If we exclude the AI accelerators, networking connectivity represented about $8 billion, and this is purely silicon, not systems, not cable nor subsystems. In fiscal 2024, we expect networking revenue to grow 30% year-on-year, driven by accelerating deployment of networking connectivity and expansion of AI accelerators in hyperscalers.”
The infrastructure software segment grew by 7% YoY to $2.0 billion and accelerated from 5% growth in the July quarter.
Due to the cyclical correction, server storage connectivity Q4 revenue declined by (17%) YoY to $1 billion. For FY2023, it grew by 11%. However, the cyclical weakness is expected to continue, and revenue is expected to decline in the mid-to-high teens for FY2024.
Broadband Q4 revenue also declined by (9%) YoY to $950 million due to the cyclical correction. The management expects the trend to continue, and for FY2024, revenue is expected to be down in the low-to-mid teens. In FY2023, broadband revenue grew by 8% YoY to $4.5 billion.
Wireless revenue declined by (3%) YoY to $2 billion. In FY2023, revenue was down (2%) YoY and the management expects revenue to be stable in FY2024. Lastly, the industrial resale revenue was flat at $236 million.
In the Core software segment, the consolidated renewal rates averaged 119%, up from 117% in the previous quarter. In the strategic accounts, it averaged 130% and was the highest in the last five quarters.
EPS:
EPS came in at $8.25 compared to $7.83 in the same period last year. The adjusted EPS came at $11.06 compared to $10.45 for the same period last year.
Analysts expect adjusted EPS to grow 0.9% YoY to $10.42 in the next quarter.
Prior to acquisition, VMware reported adjusted EPS of $1.83 in its last quarter (July 2023) as a public company.
Margins:
Gross margin for Q4 FY2023 ending Oct was 68.9% compared to 66.4% in the same period last year and 69.5% in the previous quarter.
The operating margin improved 100 bps YoY to 45.6%. The adjusted operating margin improved 20 bps YoY to 61.8%.
The company has a very strong bottom line. The net margin improved by 30 bps YoY to 37.9%. The adjusted net margin improved 90 bps YoY to 51.8% and was flat sequentially.
The adjusted EBITDA margin was 65.1% compared to 64.1% in the same period last year and 65.4% in the previous quarter.
VMWare’s EBITDA margin prior to acquisition in the July quarter was 28.6% compared to Broadcom’s 55.3%. The operating margin of VMware was 16% compared to 43.4% for Broadcom.
The adjusted EBITDA for the FY2023 ending Oct was 64.8%, and the management guide for the FY2024 is 60%, including VMware. The margin drop is mainly due to VMware's current lower margin. However, through cost-cutting initiatives like job cuts, Broadcom will aim to reach a 65% adjusted EBITDA margin for VMware. One focus area is SG&A expense, which constituted around 41% of revenue for VMware compared to around 4% for Broadcom. Hock Tan said in the earnings call, “At steady state, we’ll get to pretty close to 65% on VMware.”
The management also clarified that they are confident of achieving $8.5 billion EBITDA from VMware.
Harlan SurHarlan Sur
And then, just on my first question, are you guys still targeting $8.5 billion of EBITDA in three years on VMware?
Hock TanHock Tan
As Kirsten indicated, as we exit fiscal ‘24, we are practically at a run rate of $8.5 billion EBITDA.
Management said that the integration of VMware will take until the end of the fiscal year and require about $1 billion in transition spending. The CFO, Kirsten Spears, said in the earnings call, “During fiscal ‘24, we expect to incur about $1 billion of spend related to transitioning VMware into the new Broadcom model. This transition spending will be largely completed by the end of the fiscal year as our VMware spending run rate exits fiscal ‘24 at approximately $1.4 billion per quarter, down 40% from a year ago.”
The adjusted EPS for FY2024 is expected to grow 10.8% YoY to $46.83 and a further 19.4% YoY to $55.90 in FY2025. UBS analyst, in a research note, had earlier said that the VMware deal “to be 10% accretive to 2024 EPS and about 17% accretive to 2025.”
Cash Flow and Balance Sheet
Broadcom uses its large cash margin to frequently acquire companies.
Operating cash flow margin for Q4 FY2023 ending Oct was 51.9% compared to 53.2% in the previous quarter.
The free cash flow margin was 50.8% compared to 51.8% in the previous quarter.
The company has cash of $14.2 billion and debt of $39.2 billion. While the short-term debt is $1.6 billion, about $31.3 billion will mature after FY2028. At the end of FY2023, the company took a loan of $30.39 billion to finance the VMware merger and assumed $8.3 billion of VMware debt. While high debt is a concern, most of Broadcom’s debt has long maturities, and the company generates strong cash flows.
The company spent $15.3 billion for FY2023 in cash dividends and share repurchases. It had $7.2 billion remaining in the authorized share repurchase program.
More AI Commentary:
Vivek Arya, Bank of America analyst asked about Broadcom’s participation in the $400 billion AI accelerators market.
Hock Tan
“And we’re seeing this as we all are seeing LLM models continue to change and the face — the shape of generative AI dynamically change more and more, where training and inference are now starting to, in a way, converge and the chip designs are changing. And we are seeing that in the way we design specific custom chips for hyperscalers. That’s interesting. So that’s a very interesting opportunity for us. And as I indicated in my remarks, we see that revenue as part of networking revenue, $4 billion and networking — AI networks and going — doubling almost during 2024. Nothing new. We have said that before. And if anything else, we are reinforcing that particular guidance.”And as I indicated in my remarks, we see that revenue as part of networking revenue, $4 billion and networking — AI networks and going — doubling almost during 2024. Nothing new. We have said that before. And if anything else, we are reinforcing that particular guidance.”
Conclusion:
My takeaway is that the I/O Fund is likely to own Broadcom this year, but our goal is to enter at a lower price. Broadcom is trading above its 3-year median on PE Ratio at 40 PE compared to a 28.5 median. The current sales valuation of 15.5 PS is about double the 5-year median of 7.5 and about double the 3-year median of 8. When we go back to 2014, Broadcom has only traded above a PS Ratio of 10 one time at the height of the 2021 market. However, at a certain price, Broadcom belongs in our AI portfolio. Over the next few years, Big Tech is likely to diversify away from Nvidia’s GPUs, plus Ethernet networking continues to be upgraded for AI purposes, which means Broadcom should be on our radar.
In addition to valuation, for our purposes, we think the timing will be better once the VMware acquisition has settled as there are mixed reports from the customer perspective. Once this blows over and the restructuring is complete, it will be a more optimal time to enter Broadcom, which is richly valued at the moment.
AMD provided a solid report yet the market is struggling to find room in its valuation. Similar to Microsoft, the only time that AMD has traded this high on sales was in 2020 and again at the top 2021.
By my estimation, the selloff has nothing to do with AMD’s AI potential, or even the sequential decline in the other segments, rather it’s about timing and tech being stretched overall. We covered this in great detail in the Q1 webinar.
It’s true that AMD is trailing a GPU giant, and any guide will pale in comparison to Nvidia’s glorious 2023 performance. Yet it’s also true that AMD is coming out swinging with $3.5B in GPU revenue in the first quarter the GPUs are shipping. In a previous note, I stated: “Technically speaking, this report [Q3] was stronger than Nvidia’s was this time last year despite the GPUs shipping at the same time seasonally – NVDA had 1% QoQ growth for data center revenue in the October quarter and then declined (-6%) QoQ in the Q4 January quarter. To jog your memory, it was the guide provided in May (for the July quarter) when Nvidia had its blowout quarter.” To compare, data center growth for AMD is 38% YoY and 43% QoQ. However, Nvidia enjoyed the luxury of taking the market by surprise, and so expectations were low.
The $2 billion guide for FY2024 GPU revenue from last quarter was raised to $3.5B “plus” this quarter. Part of the issue is that analysts, independently, are creating very high expectations. The whisper number going into the report was $3 billion to $3.5 billion with some pushing for a $6 billion guide. The reason investors should separate the wheat from the chaff is that raising GPU sales by 75% in one quarter is actually very impressive (and more than enough for one quarter for our purposes). As a reminder, this is the first quarter the MI300s are shipping.
AMD will have to face outsized expectations as the company ramps, yet the 75% raise is a great start. We dissect the rest of the report for you below.
Financials:
Revenue and EPS:
Revenue of $6.28 billion was in line with management guidance of $6.1B +/- $300M. Analyst consensus was $6.14B.
The guidance for Q1 missed with $5.4B guided versus $5.78B consensus.
Adjusted EPS of $0.77 came in as expected with $0.69 adjusted EPS expected next quarter. GAAP EPS was $0.41.
Margins:
The adjusted gross margin guide expanded by one point and this is expected to continue to expand as data center increases in the product mix.
Gross margin of 47% was in line with last quarter
Adjusted gross margin of 51% and adjusted gross profits of $3.14 billion was in line with management guidance. The company guided for adjusted gross margin of 52% next quarter.
Operating margin of 6% expanded by two points and has been steadily increasing over the past few quarters.
Adjusted operating margin of 23% was in line with management guidance for adjusted operating profit of $1.41 billion.
GAAP net margin of 10.7% is a nice beginning to the bottom-line recovery for AMD, up from 5% in the previous quarter and up from 0% in the year ago quarter. Adjusted net margin was 18.6% for adjusted net income of $1.25 billion.
Regarding margins, it was indicated that the second half of the year would show improvement. Per the CEO: “I would say the headwinds side [of gross margin] continue to be in the first half where we see Embedded business not only Q1 we see sequential decline, Q2 probably are going to be sequentially flattish versus Q1. That is a headwind for us. Because it does have a very nice gross margin. But overall, we feel pretty good about the trajectory of the gross margin improvement, especially second half.”
Cash:
AMD reported operating cash flow of $381 million for a margin of 6.1%. This is lower than the previous quarter at 7% and the year ago quarter at 10%. This was due to “paying $550 million in cash taxes, primarily due to previously deferred taxes from California disaster relief efforts made available by the IRS.”
Free cash flow of $242 million for a margin of 4% was 1 point lower than the previous quarter and 4 points lower than the year ago quarter.
Product Mix:
Data Center:
The data center revenue was $2.3 billion, up 38% YoY and up 43% QoQ. The full year data center revenue was $6.5 billion, up 7% YoY. Next quarter, data center revenue is expected to be flat with “a seasonal decline in server sales offset by a strong Data Center GPU ramp.”
As stated, GPUs are expected to grow sequentially and exceed $3.5 billion in FY2024. In Q4, GPUs were $400 million of the $2.3B in data center revenue. There was an overabundance of questions about the GPU number which I’ve detailed for you below. However, here is one comment the CEO said: “And so, we worked closely with our supply chain partners to ensure that we can ship more than $3.5 billion, substantially more depending on what customer demand is as we go into the second half of the year.”
The CFO stated: “Yeah, Harsh. Let me answer your question about the MI300. Exiting Q4 2024, is it possible to get to $1.5 billion? It is possible, right, because Lisa mentioned earlier, we'll see sequential increase in each quarter and more back-end loaded in second-half and we do have a supplies more than $3.5 billion. And of course we will continue to make progress with our customers. So the math, yeah, it's possible, but right now we are really looking at focus on the execution of the current $3.5 billion plus.”
Client Segment:
The Client segment reported Q4 revenue of $1.5 billion, up 62% year-over-year yet was flat sequentially. This is lower than my notes anticipated as the guide last quarter for Client was “sequentially, Client revenue to increase.”
The CFO stated “Client is very similar to server, typically Q1 is high-single-digit to low-double-digit. That's consistent with past.”
Data center GPUs dominated the Q&A whereas this segment is incredibly important to AMD’s near-term price action (near-term is defined as this year, particularly in the second half). I’m not concerned about “if” it’ll happen rather “when” the AI PC market will become a large contributor to AMD’s revenue. Will it be this year or next? If I had attended the call, I would have asked more about this as a 75% increase in GPU revenue is plenty for one quarter when they’ve only begun to ship these units, whereas more clarity on this segment can help time AMD’s next leg up.
Moving forward, for next quarter, Client is expected to decline sequentially. This is normal for Q4 to Q1 due to seasonality. The CFO stated the following about 2024: “Directionally, for the year, we expect 2024 Data Center and the Client segment revenue to increase driven by the strengths of our product portfolio and the share gain opportunities.”
We dug up the following analyst note regarding the Client segment potentially rebounding in H2 2024: (01/23) Although PC shipments were reduced during the fourth-quarter, "we expect standard server requirements will normalize in the coming quarters, creating upside to server demand beyond the mid-single-digit growth anticipated by industry analysts," Wedbush analyst Matt Bryson wrote in a note. The anticipated recovery is in part due to the "replacement cycle" of PC's purchased during the pandemic, which is expected to occur during the second half of this year, Bryson said. "We are lifting our estimates for 2024 and 2025 given more confidence in our expectation standard server builds lift, and our belief that PC sales dip less seasonally and recover more in the second half of the year," Bryson wrote.
Gaming:
Gaming is weighing on AMD’s numbers as this segment has not bottomed yet. The segment reported $1.4 billion in revenue and was down (-17%) YoY and down (-9%) QoQ.
The guide on gaming was a lowlight: “Gaming segment sales are expected to decline sequentially, with semi-custom revenue expected to decline by a significant double-digit percentage.” This means Q4 is certainly not the bottom on gaming either. The CFO stated the same about FY2024 Gaming revenue which is that “Directionally, for the year […] Gaming segment revenue to decline by significant double-digit percentage.”
It was also stated by the CFO that gaming will decline by more than 30% in Q1: “On the Gaming side, Lisa mentioned during his — her prepared remarks is we have the latest stage of product cycle in the year five of gaming console. But at the same time, we also have inventory at the customers. So, the combination of those impact, we expect the Q1 Gaming sequential declines probably more than 30%, so hopefully that help you a little bit.”
Embedded:
Embedded reported $1.1 billion in revenue, which is a decline of (-24%) YoY and (-15%) QoQ. This is due to increased inventory. The CFO stated the following about 2024 Embedded revenue: “Directionally, for the year […] Embedded segment revenue to decline” yet later stated: “And at the same time Embedded is going through the bottoming process. We do think the second-half we will see the recovery.”
Management did highlight that Embedded drove a 25% increase in design wins to $10 billion. This will come in handy for the automotive push we want to see our portfolio participate in 2026-2028, which I also discussed on the webinar.
Earnings Call:
MI300 GPU Questions Were Plentiful:
Most of the questions on the call were fruitless attempts to get AMD to raise the 2024 GPU guide. Instead, management stuck to the $3.5 billion plus comments. Overall, it’s quite clear that $3.5 billion will become a distant memory as we move through the year but management isn’t willing to give away too much, too soon.
Joseph Moore
Great. And for my follow-up, I guess a lot of the forecasting of your business that I'm hearing is coming from supply chain and we're sort of hearing AMD is building X in Asia. I guess, how would you ask us to think about that? Are you looking at being kind of sold-out for the year and so the supply chain would be close to revenue? Are you building for the best-case scenario? Just I worry about sometimes expectations when people hear the supply chain numbers. And I'm just curious how you bridge the gap.
Lisa Su
Yeah. So, I mean, Joe, I think we updated our revenue expectations this quarter from our original number of $2 billion to $3.5 billion to try to give some bounding on some of the discussion out there. The way to think about the $3.5 billion is these are customers that we're working with, who have given us firm commitments on what they need. As you know, the lead times on these products are quite long. So, it's important to have those forecasts in early and we have a strong order book. So, that gives us good confidence to exceed the $3.5 billion. From a supply chain standpoint, our goal is always to build more supply we — and so, from that standpoint, we have also worked with our supply chain partners and secured significant capacity.Think about it as first half capacity is tight and more comes on in the second half of the year, but we've certainly made more progress there. So, we do have more supply, and we're going to continue to work with our customers on their deployments and we'll update that number as we go through the year.”
Here was another good discussion on the MI300 and the trajectory investors can expect in 2024:
Toshiya Hari
Hi. Thank you for taking the question. I had one on the MI300 as well, Lisa. I guess, first of all, how should we think about the quarterly trajectory beyond Q1? You talked about Q1 being up sequentially. Is it fair to assume kind of a straight line as we progress through the balance of the year? Or is it more second half skewed? How should we think about that? And I guess more importantly, some of the cloud potential customers that have yet to officially sign-up for or sign-off on the MI300. I guess what's the sticking point? Is it just a function of time and you just need a little bit more time to go back-and-forth and tweak things or is there a software kind of concern? I guess what's holding them back at this point?
Lisa Su
Yeah, Toshiya, thanks for the question. So, first on the MI300 trajectory. I think you would expect that revenue should increase every quarter from now through sort of the end of the year, but it will be a bit more second half weighted and part of that is just customers as they're finishing up their qualifications in their lines as well as sort of how our supply chain is ramping. So, yes, it should increase each quarter, but be a bit more second half weighted. And then, to your comment about customers, look, we are engaged with all of sort of the large customers. These are all folks that know what's really well, given our deep relationships in EPYC. I think people just have different adoption cycles as they consider what they're trying to do in their roadmap. But I view this as still very, very early innings for us in this space. And I think the question was asked earlier. I think the key is this is not just about MI300 conversation. But it really is about sort of our long-term multi-generational roadmap. And so, that's the context on which we're working with our largest customers, as well as, as you know, there's a lot of demand coming from folks that are more AI centric and not necessarily typical cloud customers, but more enterprise or let's call it AI-specific companies that we're also very well engaged with.”
More on Embedded and Gaming:
The CFO provided the following color on embedded and gaming:
Jean Hu
Yeah, Aaron, I'll give you some color about Client seasonality and others. So, Client is very similar to server, typically Q1 is high-single-digit to low-double-digit. That's consistent with past. On the Embedded side, it's very consistent with what we said in the past and the consistent with what you see in the industry's Embedded business is going through a bottoming process, and we think Q1, it will have a low-double-digit sequential decline. That's Embedded. On the Gaming side, Lisa mentioned during his — her prepared remarks is we have the latest stage of product cycle in the year five of gaming console. But at the same time, we also have inventory at the customers. So, the combination of those impact, we expect the Q1 Gaming sequential declines probably more than 30%, so hopefully that help you a little bit.
Conclusion:
AMD has been kind to investors over the past year with 200% gains since Jan 1, 2023 – and the kicker is that AI revenue was officially realized in Q4 for the first time, at a mere $400 million. The market is correct to anticipate that AMD will be a major player in AI. How long until AMD becomes officially number two? Right now, Broadcom is expected to report $7 billion in AI revenue for 2024. I think AMD will surpass Broadcom very quickly.
From there, it will be a battle – a fun one for us to observe as we own both NVDA and AMD. Readers on our site know that I call AMD the “Dark Horse” because it’s frequently underestimated. The stock sold off in July when AMD held their AI event to officially announce the MI300s. Today, the stock is selling off in the quarter the company is shipping them. Clearly, the market sees potential, hence the stock performance. Yet, AMD investors should be prepared for periods of doubt, similar to when AMD emerged as a leading CPU design company. We are looking to be entering one of those periods now.
With that said, tech has a way of becoming stretched, especially following a historic year for the Nasdaq (btw, 2023 was the leading year for gains in over 20 years since dot-com era, edging out 2009), and thus we have not been buyers of AMD as of late. In the near-term, following our technical analysis and broad market plan is going to be as equally important as following our stock analysis.
On that note, this Thursday, Pro Members are invited to attend a special 1-hour webinar with Knox Ridley as he goes through what investors can expect for tech stocks in Q1 and the risks present in the market. We encourage our Members to attend the webinar or watch the replay as the I/O Fund continues to deeply strategize for 2024. Keep an eye out for details via your inbox tomorrow.
Housekeeping: Mailchimp was down system-wide for a period of time yesterday, you may have received emails at sporadic times once MC was operable again! We seem to be back up and running just fine now.
Demand is too high for GPUs with long lead times, this is an excellent window of time for a competitor to enter the picture.
The MI300s should be able to compete on performance once the GPUs are benchmarked as AMD’s GPUs power the world’s largest supercomputers.
AMD is exceptional at undercutting on price. This is primarily how AMD overtook Intel coupled with a better design (the Zen 2 architecture)
AMD’s designs are excellent at improving power efficiency. Power efficiency is important for total cost of ownership. Not only will AMD’s GPUs likely be cheaper (no confirmation on pricing just yet) but they will also cost less to own over a four-year life span.
Hyperscalers will support competition to Nvidia. You can think of Nvidia as more of a frenemy to Big Tech. This is due to pricing power, CUDA being closed source, and also now Nvidia will be competing with Big Tech in some areas. For example, Omniverse will compete with Meta’s metaverse ambitions. For the MI300 release, AMD is primarily focused on hyperscalers with the CDNA GPUs and not consumer-level RDNA GPUs.
Here are the challenges AMD faces:
Lacks a popular software platform and CUDA competitor. AMD’s recently released software platform ROCM is promising but is no CUDA.
AMD is later to market on AI acceleration in terms of GPUs. Although AMD has accomplished what is nearly impossible by being a second-place contender that crushed first-place Intel, the reality is that being in second place is a major obstacle.
On that note, the company has its hands full competing against Intel on CPUs. It will now go up against Nvidia on GPUs. Lisa Su is one of the best CEOs in the history of the tech industry, but can she and her team take on both at the same time?
Revenue and EPS:
AMD bottomed in June with (-18%) growth for revenue of $5.4 billion. The company reported growth of 4% YoY in the June quarter for revenue of $5.8 billion.
In the upcoming December quarter, AMD is expected to report $6.14 billion for growth of 9.64%. This is within range of what management guided last quarter:
“For the fourth quarter of 2023, AMD expects revenue to be approximately $6.1 billion, plus or minus $300 million. At the mid-point of the revenue range, this represents year-over-year growth of 9% and sequential growth of 5.2%.”
Next quarter, AMD is expected to report revenue of $5.78B for growth of 7.97%.
Note, revenue has been revised downward recently as we had estimates at 15.65% following the last earnings report. As a reminder, AMD’s stock was rocky last earnings report as it sold off based on the print reflecting a Q4 miss, the stock later recouped when it was clarified the miss was not coming from the data center, rather the gaming sector and the embedded segment are weighing on revenue.
AMD is expected to double adjusted EPS over the next seven quarters with estimates at $0.77 this quarter and $0.68 next quarter.
Margins:
AMD guided for gross margin of 51.5% this quarter for adjusted gross profit of $3.14B. This is in line with previous quarters.
Adjusted operating margin was guided to be 23% with $1.74B in opex, which leaves adjusted operating income of $1.4 billion. This will be higher than the previous three quarters, and flat YoY.
Last quarter, adjusted net margin was 5% for adjusted net income of $1.13B. This has been trending upward and reflects the rebound in revenue.
Cash:
Operating cash flow last quarter was $421 million for a 7% margin. Free cash flow was $297 million for a 5% margin. The company has cash and short term investments of $5.8 billion and debt of $2.47 billion.
Revenue Segments:
Data Center:
Last quarter, data center reported growth of $1.6 billion. From there, management has guided data center and the Client segment to be “up by strong double-digit percentage. Sequentially, expect Data Center segment to grow by strong double-digit percentage.”
As stated in our previous earnings write-up, it was confirmed the Q4 data center will grow 50% QoQ for an estimated $2.4 billion. To jog your memory, for Q4, the $400 million in GPUs are mainly from the El Capitan supercomputer.
Per the opening comments last quarter: “Based on the rapid progress we are making with our AI road map execution and purchase commitments from cloud customers, we now expect Data Center GPU revenue to be approximately $400 million in the fourth quarter and exceed $2 billion in 2024 as revenue ramps throughout the year. This growth would make MI300 the fastest product to ramp to $1 billion in sales in AMD history.”
As you’ll recall, this comment was important because it means commercial customers will contribute in 2024. This is primarily what the market will be watching – not only data center growth but any color provided for GPUs. Right now, $2 billion is the benchmark – let’s see if AMD can give us a higher number for FY2024.
Client Segment:
Last quarter, the Client segment reported revenue of $1.5 billion, which was up 42% YoY and 46% QoQ driven by Ryzen mobile processor sales. Management has stated the Client segment is also expected to fully rebound by a “strong double-digit percentage.” Also, management stated that sequentially, “client segment revenue to increase.”
Gaming Segment:
Last quarter, the gaming segment reported (-8%) YoY and was down (-5%) QoQ due “to a decline in semi-custom revenue partially offset by an increase in Radeon GPU sales.” This quarter, gaming is expected “to decline by double-digit percentage”
Embedded Revenue:
Last quarter, Embedded reported (-5%) YoY and was down (-15%) QoQ. This was due to “an inventory correction at customers in several end markets.” The guide for this quarter is that the Embedded segment will decline “due to additional softening of demand in the embedded market. Sequentially, the embedded segment is expected to decline by double-digit percentage.”
More comments on the $2B
Toshiya Hari:
[…] My first one is on the Data Center GPU business. You talked about '24 revenue potentially exceeding $2 billion. I was hoping you could provide a little bit more color. What percentage of this is AI versus supercomputing or other applications?
Lisa Su:
Your question as to how the revenue evolves, so the way to think about it is, in the fourth quarter, we said revenue would be approximately $400 million, and that's mostly HPC with some — the start of our AI ramp. And then as we go into the first quarter, we actually expect revenue to be approximately similar in that $400 million range. And that will be mostly AI so with a very small piece being HPC. And as we go through 2024, we would expect revenue to continue to ramp quarterly, and again, it will be mostly AI.
The CEO let a comment slip that it would be “greater than $2 billion” for next year, and given how careful this management team is, I do think it’s important to note here:
Lisa Su:
Sure, Aaron. So, we've been planning the supply chain for the last year and we're always planning for success. So, certainly, for the current forecast of greater than $2 billion, we have adequate supply. But we have also planned for a supply chain forecast that could be significantly higher than that, and we would continue to work with customers to build that out.
For our purposes, as far as where this can go, the management team goes back to quoting 50% CAGR in this segment over the next few years: “So, I think we are big believers in the strength of the market. We previously said we believe that the compound annual growth rate could be 50% over the next three or four years.”
What is a bit unfortunate is that the gaming segment and embedded is weak, and these two segments overshadow AMD continuing to take market share on the CPU data center, and the company’s highly anticipated answer to Nvidia’s H100s.
Per the CFO’s opening remarks: “In the fourth quarter, we expect to benefit from strong Data Center and Client momentum, driven by MI300 AI accelerated ramp and the strength of our high-performance leadership Zen 4 family of products despite lower sales in the Gaming segment and additional softening of demand in the embedded market.”
What is quite fortunate, however, is that some of AMD’s AI story is undercover, which is the PC and mobile market. There’s a bonus waiting for AMD investors as the AI story plays out, and while data center GPUs take up all of the attention, there is another powerful AI trend that is silently building strength in the background. I mentioned this in a previous write-up: “When discussing AMD’s AI opportunity, it is vitally important that we not lose sight of the opportunity AMD will have to expand its AI portfolio to the Client Segment.” This has already begun with the Ryzen AI on-chip accelerator contributing to the rebound in the client segment this quarter.
From the CEO in the last earnings report: “What I'm most excited about in PCs is actually the AI PC. I think the AI PC opportunity is an opportunity to redefine what PCs are in terms of productivity tool and really sort of operating on sort of user data. And so, I think we're at the beginning of a wave there. We're investing heavily in Ryzen AI and the opportunity to really broaden sort of the AI capabilities of PCs going forward.”
Our post-earnings report will hit your inboxes tomorrow night.
Supermicro has produced a neck-breaking 1-month return of 84% in four weeks, if we assume it opens at $547 tomorrow. This is exceptional as QQQs are only up 4%. As a reminder, it’s been one of the best performing stocks for two years in a row. Supermicro personifies the saying “winners keep winning.”
Last quarter, we had stated that Super Micro’s guidance appeared to be conservative: “By far, the most important question on the call was if management was being conservative in their fiscal year guide OR are there challenges ahead for the March and June quarter? Management provided a straightforward answer that Q4 and the fiscal year guides are conservative, implying that we have chance for a beat again in the near-term.”
The CEO further repeated the word “conservative” a few times last quarter. We like this as it creates trust with the management team, especially as SMCI was being doubted in a bout of weakness for the stock during the last report, being down (-12.7%) that month. We were able to rely on management’s comments irrespective of what the market was doing, and this is our preferred path.
The word “conservative” was used again, and an analyst attempted to model how SMCI could exit the year. If the analyst is correct in modeling the opportunity, then there’s roughly 60% room in the stock’s valuation. Each investor will need to decide for themselves how to interpret the Q&A. However, that was my takeaway and it’s likely we take a shot at adding to our position tomorrow.
Before I get too caught up in describing the revenue beat, I do want to say that cash was a blemish for this report and it’s important to file a note that cash flow needs to be watched. We detail this below under the Cash section.
Revenue and EPS:
Similar to Nvidia’s streak last year, analysts cannot keep up with Super Micro this month. As soon as the analyst consensus is revised, Super Micro delivers again.
Here is what that has looked like:
Originally, the current Q2 quarter consensus was $2.49B at the midpoint with analysts revising it to $3.26B. Super Micro beat this with actual revenue of $3.7B for growth of 103% YoY and 73% QoQ.
For the forward Q3 quarter, consensus was $2.99B and Super Micro has now guided for revenue of $3.7B to $4.1B, representing YoY growth of 204% at the midpoint.
Fiscal year estimates were revised upward to $11.5B on January 22nd, yet Super Micro guided for $14.3 billion to $14.7 billion, at the midpoint. This represents growth of 103.6% for the fiscal year. In the opening remarks, the CEO referred to “continued strength in the second half of 2024.”
GAAP EPS for December was $5.10 compared to the analysts’ consensus estimates of $4.90. This is nearly 80% higher sequentially with $2.85 GAAP EPS in the previous quarter and is 62% growth from the year ago quarter. Adjusted EPS was $5.59 for similar YoY and QoQ growth.
Looking forward, next quarter’s GAAP EPS is expected to be between $4.79 and $5.64 for over 240% growth from the year ago quarter. Adjusted EPS of $5.60 at the midpoint will see similar YoY growth.
Margins:
The margins on SMCI are thinner than most semiconductors, and this leads to questions on the earnings call. We detailed you this quarter’s questions about the margins for you below.
Gross margin was 15.4% for gross profit of $564.4 million
Operating margin was 10.1% for operating profit of $371.5 million and adjusted OPM was 11.3%
Net margin was 8% for net profits of $295 million. Adjusted net margin was 9%.
Cash:
Operating cash flow was (-$595) million for a negative margin of (-16.2%). This compares to +12.8% last quarter and +9% in the year ago quarter. Per the CFO: “Cash flow used in operations for Q2 was $595 million compared to cash flow generated by operations of $271 million during the previous quarter. Strong profitability and higher level of accounts payable was offset by higher inventory and accounts receivable due to build plans for Q3 and the timing of shipments during Q2.”
Free cash flow was (-$610) million for a margin of (-16.6%) compared to +12.7% last quarter and +8.4% in the year ago quarter.
The company has $726 million in cash on its balance sheet and $376 million in debt. This is up from $543 million in cash last quarter and net debt of $146 million. This results in net cash position of $350 million compared to a net cash position of $397 million last quarter.
The company was able to increase its cash through an equity offering: “During the quarter, we executed an equity offering and raised approximately $583 million in net proceeds after underwriting discounts and other issuance costs from the sale of 2.3 million shares at a price of $262 per share. The proceeds will be used to strengthen our working capital, enable continued investments in R&D and expand global capacity to fulfill strong demand for our leading platforms.”
Key Metrics:
Revenue Segments:
OEM Appliance and Large Data Center segment was $2.15 billion for 59% of revenue. This was up 175% YoY and 83% QoQ. Per management: “Two existing CSP/large data center customers represented 26% and 11% of total revenues for Q2"
Organic (Enterprise & Channel), AI/ML was $1.48 billion for 40% of revenue, up 55% YoY and up 62% QoQ. This was driven by “enterprise AI and CPU upgrade programs.”
5G, Telco and Edge/IoT was 1% of revenue for $35 million.
Revenue Mix:
Server and storage systems were $3.4 billion in revenue for growth of 107% YoY and was 94% of Q2 revenue
Subsystems and Accessories were $229 million, up 61% YoY and was 6% of Q2 revenue.
Inventory:
Inventory days decreased to 67 days compared to 91 days in the previous quarter. This is aligned with management’s comments on tight supply.
Geography:
All revenues were up by a wide margin QoQ.
United States was 71% of revenue, slightly down from 76% last quarter. Revenues increased 139% YoY.
Asia was 18% of revenue, up from 11% last quarter. Revenues increased 98% YoY.
Europe was 8%, down marginally from 9% last quarter. Revenues decreased 8% YoY.
ROW was 3% compared to 4% last quarter. Revenues increased 67% YoY.
Earnings Call Q&A
“Conservative” Referenced Again
My ears perked up when I heard the following from question:
Samik Chatterjee
Got it. And then just this more near-term question when I look at the revenue guide for 3Q and 4Q. Is a step-up here in revenue of about sort of call it $0.5 billion a bit less going from 2Q to 3Q. And then a bigger step-up to get to the mid-point to be annual guide into 4Q, how much of that is driven by just being a bit more cautious about when supply comes in and pushing that the revenue guide a bit more to the 4Q or is that really what the visibility currently of supply of just trying to get sort of what's driving the cadence from 2Q to 3Q to 4Q. And the guide that you provided. Thank you.
David Weigand
Yeah, so Samik, we have a very large and growing backlog, which grew again this quarter. And so really as Charles mentioned earlier, our only constraint is supply. However, the good news is, supply is improving. And so to your point, we have to be somewhat conservative, because we are constrained still by supply.
Regarding how far and how quick the company can grow, it doesn’t hurt that Super Micro is preparing supply to break the $25 billion revenue mark:
“Today, our production utilization rate is about 65% across our USA, Netherlands and Taiwan facilities, and they are quickly filling. To address this immediate capacity challenge, we are adding two new production facilities and warehouses near our Silicon Valley headquarter, which will be operating in a few months. The new Malaysia facility will focus on expanding our building blocks with lower costs and increased volume, while other new facility will support our annual revenue capacity above $25 billion.”
Later in the call, an analyst modeled SMCI ending the year with a $5.6 billion quarter. The CEO seemed to agree. I’m quoting it here in full so each investor can form their own takeaway.
“Ananda Baruah
Yeah, good afternoon guys. And thanks for taking the question. Appreciate it. Congrats on the really solid execution. Yeah, congrats on that. I guess, two if I could, Charles. And maybe a clarification, I did some math on the 15,000 racks per month. And they came up with — I guess $5.6 billion a quarter, let's call it 5.5%. I guess, plus or minus but that came to $5.6 billion, is that kind of accurate and I guess the question is — if it stood at midyear, you're talking about getting to that point. Is that the kind of run-rate opportunity that we can be thinking about quarterly and not like a guidance, but like an opportunity when you get into sort of the back half of the calendar year. Just wanted to make sure that we're interpreting that kind of accurately and then I have a quick follow-up. Thanks.
Charles Liang
Yeah, again, we see, we recommend green computing everywhere. That's why wherever, whenever we can help a customer at base, we will. That's why we have been building really large-scale capacity for liquid cooling and other green computing solutions. So yes, that capacity will be huge, but its capacity there, when customer need, we are ready. And indeed our facility also very flexible. Lots of facility can support liquid cooling and air cooler, or combination cooling. So yes, we have a huge capacity ready for growth, but not necessarily all for liquid cooling, they support air cooler or combination hybrid cooling as well.”
This is based on comments from the CEO that: “By this June quarter, we will have high volume, dedicated capacity for manufacturing 100 kilowatt to 120 kilowatt racks with liquid-cooling capabilities, providing DLC, direct liquid cooling racks capacity up to 1,500 racks per month and our total rack production capacity will be up to 5,000 racks per month by then.” This means SMCI will exit the calendar year capable of 14,000 racks.
Questions on the Margins
Notably, Super Micro has weaker margins than was Wall Street prefers. There is always a question about margins on the call. There was a large revenue beat that did not flow through to a higher gross margin or operating margin, and in this Q&A session, the CFO stated: “And so, at this time we are we are growing really quickly. And in order to do that and in order to take market share, we will take opportunities by being more competitive on pricing.” The CEO followed up with: “The good thing is that when we continue to grow our economies of scale, our operation margin indeed will be still able to keep in healthy position.”
Here was a more specific question:
Jon Tanwanteng
Got it, and to ask one of the questions that's been mentioned in a different way, is there a gross margin floor as you pursue this share gain. And when do you see a possible inflection? I'm just wondering what is the limit. When you go in terms of gaining share versus the margin that you generate?
David Weigand
Sure, so we set out a target back in March of 2021 of 14% to 17%. But that's and we've actually done pretty well against that target.
The Mystery 11% Customer
With Intel clearly stumbling (again!), I think the 11% Mystery Customer could be AMD. This is pure speculation, and one we don’t have to wait too long to find out given AMD reports tomorrow night.
Aaron Rakers
Yeah, thanks for taking the questions and also great results. Just curious, when you talk about your customer concentration and the diversity of the business, when you talk about 26% and 11% of your revenue coming from two customers, are those the same customers, like last quarter I think you had a customer that was 25%, or are you seeing these customers kind of bounce around. I guess the simple question is just, is that the same customer 26% and 25%?
David Weigand
So, Aaron. The 26% customer is the same customer. But the 11% customer is not a new customer, and it's a longer-term customer. But first time in 11% and to your point, yes, we do see a bouncing in and out, and we're very happy. Anytime they do bounce about by the way.
Liquid Cooling Reaches Inflection with Nvidia’s B100s
Super Micro is primarily air cooled right now, yet liquid cooling is growing. Per the CEO in the opening remarks: “By this June quarter, we will have high volume, dedicated capacity for manufacturing 100 kilowatt to 120 kilowatt racks with liquid-cooling capabilities, providing DLC, direct liquid cooling racks capacity up to 1,500 racks per month and our total rack production capacity will be up to 5,000 racks per month by then.” To read more on liquid cooling, reference our previous Super Micro analysis here.Super Micro analysis here.
Here was another comment from the CEO regarding liquid cooling reaching an inflection point with Nvidia’s upcoming release of the B100s:
Quinn Bolton
Okay, got it and then as Charles said, question on liquid cooling. Just as you look forward, you guys are ready, it sounds like the infrastructure may still need some improvements, but I guess as you look at data center customers CSPs that are looking to deploy liquid cooling. Is that sort of does that include current-generation sort of 700-watt, GPUs are, is it really the next generation, the B100s and sort of the 1,000 watt GPU class that really drives the adoption at your CSP customers, drives that need for liquid cooling.
Charles Liang
You are right. In these current 600 watt, 700, watt module, people can still take care very well, we say air conditioning. And that's why people still are comfortable with our traditional air cooler. But when that system grows to 1,000 or even 1,000 watt per module, yes. I mean — I think cooling becomes even much more critical. So, by that time, I believe. Most of the data center will have facility ready for that. So we are very optimistic and very patient to continue to improve our quality, especially that reliability and easy for maintenance. So when customers are ready, we can dwell quickly to support them.
Conclusion:
If the analyst’s model of the opportunity is correct, then Supermicro’s valuation just became quite attractive. The market cap is 27.5 billion at market close, and the run rate would be $22.4 billion in this case, leading to a 1.2 P/S on a company that was trading at a 1.9 P/S, as of market close if we use the previous December quarter estimate of $3.68 billion. This could translate to 60% room in the valuation. For our risk profile purposes, we think it’s worth a shot to add to our position tomorrow and will keep you posted.
Micron announced its fiscal Q1 earnings on December 20 after pre-announcing results on November 28. Micron beat on both the top and bottom lines, showed solid improvement in operating cash flow as margins continue to recover, and guided its fiscal Q2 revenue nearly 5% higher than consensus. Shares rose 8.6% following the report, Micron’s largest one-day gain since March 2022, as it offered further evidence that the memory market recovery may be underway with its first YoY revenue increase in six quarters.
With Micron, we need two things to happen. The first, we is want Micron to reach a bottom fundamentally as it’s been in a steep cyclical downturn. We believe the fundamental recovery is squarely underway, and we’ve included the evidence below plus a few charts that best illustrate this on a YoY and QoQ basis.
Secondly, we want the right entry on Micron which is more challenging than usual at a fundamental bottom given semiconductors had a banner year in 2023. To illustrate, semiconductor ETFs had a banner year in 2023 with SMH +72% and SOX +65% compared to QQQ +54%, this is even with QQQ greatly benefiting from the Mag 7’s brilliant year. Thus, semis are looking exhausted. Micron was up 70% in 2023 despite a steep memory trough; compare this to Nvidia’s fundamental bottom in October of 2022 when the stock was down (-60%) YTD. It’s easy to see why an entry is tricker with Micron in this case.
Micron, like a few other names, appear to have room for one more swing higher, but after this, we believe it is a reasonable probability that we get MU at a lower price. For this reason, the bulk of our targeted allocation is on the side lines, waiting for lower targets. Should AI drive timing that is quicker than getting a lower valuation, then we will buy a breakout. Overall, the bigger plan is to layer-in, which we discuss in more detail below.
For more information on Micron’s thesis and the importance of HBM, reference our analysis Micron: AI Offers a Multifaceted Secular Growth Tailwind, where we highlighted a week prior to earnings that AI offers Micron a secular growth tailwind. Demand for HBM3e chips is expected to surge, a critical component to support faster iterations of AI GPUs from Nvidia and AMD. We’ve also discussed HBM in a previous AMD write-up and Lam Research write-up.
In the earnings call, CEO Sanjay Mehrotra talked up the AI tailwinds, saying “AI has emerged as a significant secular driver that will further bolster the industry towards record revenue TAM in 2025 and drive growth for years to come,”AI has emerged as a significant secular driver that will further bolster the industry towards record revenue TAM in 2025 and drive growth for years to come,” with Micron “well positioned to capitalize on the immense opportunities artificial intelligence is fueling across end markets.”
Revenue and EPS:
There is strong evidence the top line and bottom line has bottomed for Micron. In the most recent report, revenue of $4.73 billion beat estimates by 2% representing YoY growth of 16% and QoQ growth of 18%. GAAP EPS of ($0.95) beat estimates by $0.06, or 6%. This compares to GAAP EPS of ($0.04) in the year ago quarter and ($1.07) last quarter.
Margins:
Margins have gone through a steep cyclical low, with an apparent rebound per last quarter. The GAAP gross margin was (0.7%) in Q1, compared to 21.9% in the year ago quarter. This marked a 1010 bp QoQ improvement from (10.8%) in Q4.
GAAP operating margin was (23.9%) in Q1, compared to (5.1%) in the year ago quarter. This represented a 1280 bp QoQ improvement from (36.7%) in Q4.
GAAP net margin was (26.1%), compared to (4.8%) in the year ago quarter. This represented a 960 bp QoQ improvement from (35.7%) in Q4.
Cash and Debt:
Micron’s operating cash flow has bottomed in a major way with Q1 at $1.40 billion, representing YoY growth of 48.6% and QoQ growth of 462.7%.
Adjusted free cash flow in Q1 was ($0.33 billion), an improvement from ($1.53 billion) in the year ago quarter and ($0.76 billion) in the prior quarter. The negative adjusted FCF was driven primarily by $1.8 billion expenditures in plant, property and equipment.
Cash, marketable securities, and restricted cash totaled $9.84 billion.
Debt totaled $13.33 billion. Net debt of $3.56B in 2023 is expected to be slightly worse in 2024 due to capex before it resolves entirely in 2025 due to strong cash flows.
Revenue Mix:
According to Gartner, DRAM will grow by “as much as 87 percent” in 2024. NAND flash memory is projected to decline by (-32.9%) and then “bounce back to grow by about 60 percent” in 2024. Micron’s report shows evidence that DRAM has rebounded while NAND is slower to recover.
Per our AMD write-up, memory is where AI acceleration is fiercely competing in 2024, and this will be reflected in Micron’s DRAM segment. On a QoQ basis, DRAM revenue increased 24.4% QoQ to $3.43 billion. This represents growth of 21.1% YoY.revenue increased 24.4% QoQ to $3.43 billion. This represents growth of 21.1% YoY.
This is where the AI bull thesis is centered. To put it into context, the wafer fab equipment supplier Lam has stated that “AI servers use 8X DRAM and 3X NAND compared to an enterprise class server.” We also stated in the Micron deep dive that: “HBM3 will see a surge in demand in 2023 and 2024, rising to 39% share this year and approaching 60% share in 2024 to become the dominant architecture.”
NAND revenue increased 11.5% YoY and 2.1% QoQ to $1.23 billion. NAND revenue accounted for 26% of revenue down 400 basis points. Higher average sales prices (ASP) offset lower volume of sales in NAND. An interesting note from the call was that Micron nearly tripled its market share in data center storage SSDs from 11% market share at the midpoint, up from 4% at the midpoint.
The mix of ¾ DRAM and ¼ NAND is typical for Micron
Key Metrics from Business Units:
Compute and Networking Business Unit (CNBU) revenue increased 45% sequentially to $1.74 billion, as data center and client shipments strengthened in Q1. The CEO pointed toward AI driving this strong QoQ growth: “AI-related shipments increased in the data center market and normalized inventory at client customers enabled bit shipment growth.”
Compute and networking is where we want to see more recovery as the segment was as high as $3.8B in Q4 of 2021 and $3.9B in Q3 of 2022, Micron’s cyclical peak.
Micron’s mobile business unit is also showing signs of recovery at $1.3 billion, up 7% QoQ. This is double the revenue from a year ago and is signaling a recovery for mobile. The segment was up 48% QoQ last quarter yet down (-20%). Per the CEO: “Mobile revenue continued to show strength as customer inventories normalized and smartphone units and average memory and storage capacity growth at customers drove demand.”
For reference with Micron’s peak quarters, the mobile segment was at $1.8B in Q4 of 2021 and $1.97B in Q3 of 2022.
Embedded has also bottomed (likely) at 21% QoQ growth to $1.03 billion. This compares to $1.36B in Q4 of 2021 and $1.4B in Q3 of 2022.
Storage (SBU) was down (-12%) QoQ to $653 million. This compares to $1.36B in Q4 of 2021.
Q2 Outlook:
Micron guided Q2 revenues of $5.30 billion, +/- $200 million, above the consensus estimate of $5.05 billion and representing YoY growth of 43.5% and QoQ growth of 12.1%.
Q2 is expected to start a 6-quarter streak of quarterly YoY growth above 40%, lasting through Q3 of fiscal 2024.streak of quarterly YoY growth above 40%, lasting through Q3 of fiscal 2024.
GAAP gross margin was guided at 12.0%, +/- 1.5%, pointing to a 1270 bp QoQ improvement at midpoint – quite a rapid recovery from the (10.8%) gross margin reported in fiscal Q4. Micron says it expects this improvement to stem from “sequential price increases and reduced impact from underutilization.” Pricing is expected to be the main driver of this margin improvement.
Non-GAAP EPS was guided ($0.28) +/- $0.07, far above consensus of -$0.61. GAAP EPS was guided at ($0.45) +/- $0.07, a solid improvement from Q1’s ($0.95) print. The guide alludes to a quicker shift back to profitability – Micron was previously expected to shift back to GAAP profitability in fiscal Q4, but that expectation has now shifted one quarter forward, with analysts expecting Micron to report GAAP EPS of $0.01 in fiscal Q3. However, it’s such a fine line that GAAP profitability is not a given in fiscal Q3, especially if gross margins come in below expectations in Q2.
Earnings Call:
AI Commentary incl Nvidia partnership on H200:
Generative AI’s influence on the memory market and on Micron was increasingly clear during the earnings call. Mehrotra explained that “Generative AI use cases are expanding from the data center to the edge, with several recent announcements of AI-enabled PCs, smartphones with on-device AI capabilities, as well as embedded AI in the auto and industrial end markets. … We see a rapid evolution in our customer product roadmaps enabling and leveraging this AI market expansion, which in turn is driving higher capacity, lower power, and increased performance requirements for memory and storage. We expect to increasingly benefit from content growth as these trends in AI gain momentum.”
Mehrotra also shed more light on Micron’s progress in HBM3e development – Micron is“in the final stages of qualifying our industry-leading HBM3E to be used in NVIDIA's next generation Grace Hopper GH200 and H200 platforms.”
Micron also is “on track to begin our HBM3E volume production ramp in early calendar 2024 and to generate several hundred millions of dollars of HBM revenue in fiscal 2024. We expect continued HBM revenue growth in 2025, and we continue to expect that our HBM market share will match our overall DRAM bit share some time in calendar 2025.”
Micron is estimated to have held around 21.5% to 22.8% DRAM market share in calendar Q3, with bit shipment share likely in the same low to mid-20% range – this implies Micron is aiming to capture 20% to 25% of the HBM market in 2025.
HBM revenues are projected to rise as much as 127% YoY in 2024 to $8.9 billion, per TrendForce, potentially reaching more than $15 billion assuming 70% growth in 2025. This equates to a >$3 billion revenue opportunity if Micron can indeed capture more than 20% of the market. HBM’s CAGR is expected to be more than 50% over the next few years.
Higher Prices; Tighter Supply
Aiding Micron’s recovery is improvement in the supply-demand balance, which is resulting in higher prices. The CEO pointed toward higher prices as the reason for the Q2 guidance beat: “The current pricing trajectory has improved our financial outlook for the second quarter and full fiscal year.”
The key to Micron’s potential for a beat for FY2024 is also pricing: “Our leading-edge DRAM and NAND nodes are oversubscribed for the full year. Consequently, we expect prices to increase through calendar 2024, driving improvements in our financial performance.”
Notably, per management, pricing is below industry average right now, yet we are cautiously optimistic it will lead to a positive surprise or two this year for Micron. “We have driven a strong inflection in industry pricing this calendar quarter, which will allow us to benefit from higher prices earlier in our fiscal year compared to our prior plans.”
Pricing is being boosted by tight supply. Because this dynamic is key to Micron’s ability to beat/raise this year, I’m inserting the full quote as it’s a lucid explanation of why 2024 and 2025 could be outliers for memory suppliers, especially those specializing in smaller nodes.
“In last quarter's earnings call, we communicated that we strategically diverted underutilized equipment toward ramping new technology nodes, which will help us increase leading-edge production in a capital-efficient manner. Since the number of wafer processing steps is higher for leading-edge nodes, this approach of diverting underutilized tools to the leading edge meaningfully reduces our overall wafer capacity. Thus, underutilization in our fabs early this fiscal year transitions to structurally lower wafer capacity at higher utilization rates as we move through the fiscal year. Reports indicate that this redeployment of underutilized tools at the leading edge is an industry-wide practice that is likely to constrain industry supply in 2024.
Taking all these factors into account, Micron's bit supply growth in fiscal 2024 is planned to be well below demand growth for both DRAM and NAND, and we expect to decrease our days of inventory in fiscal year 2024.We expect calendar 2024 industry supply to be below demand for both DRAM and NAND, which will result in a contraction of industry inventory levels.”
COGS Increasing
One item to watch as this recovery unfolds is cost of goods sold (COGS). In Q1, COGS increased 49% YoY, the largest YoY jump since 2014. An increased emphasis on HBM3e is a driving factor in COGS rising rapidly, as HBM3e development is more costly (Micron says HBM is “consuming more than 2 times the wafer supply as D5 to produce a given number of bits”) but likely comes with a much higher ASP.
For 2024 and 2025, this may be alleviated by tight supply and higher prices. Per the CEO: “[…] we have tightness on our leading-edge nodes. They are already in short supply and inventories will continue to improve for us. And all of this results in overall healthy dynamics for pricing improvements, profitability improvements, and revenue opportunity growth in the backdrop of demand drivers, AI being a dominant demand driver across the end markets.”
However, eventually, inventory may build and the margins will be something to watch as we go along in the long-term. This won’t be a problem in 2024 or 2025 per management comments and analyst consensus.
Risks:
The predominant risk for Micron is that it’s in third place behind Samsung and SK Hynix. Will AI help propel MU to a higher market share and/or will the emphasis to use fabs on United States soil be a tailwind for Micron? Those are the questions that have to be answered – the rest is quite bullish in terms of product story, what remains is competitive positioning, with Micron up against South Korea’s stronghold in memory.
Per the most recent earnings call, Micron is prepared to compete during the critical moment for memory to enable higher performance and lower energy AI acceleration: “Micron is addressing these exciting opportunities brought on by the proliferation of AI with an industry-leading portfolio of data center solutions, including HBM3E, D5, several types of high-capacity server memory modules, LPDRAM, and data center SSDs. We have received very positive customer feedback on our HBM3E, which has approximately 10% better performance and about 30% lower power consumption compared to competitive offerings of HBM3E.”
To remain further competitive, Micron has also released advanced DDR5 DRAM for CPUs based built on a 1-beta node and a 32Gb monolithic die. This is considered the world’s fastest and lowest latency 128 GB high-capacity modules. Per the earnings report: “Additionally, leading CPU vendors have confirmed validation support for our monolithic-die-based 128GB modules on existing platforms released in 2022 and 2023 as well as upcoming new platforms. This ensures that our offering has a significant TAM that we can address immediately. We expect volume production to start next quarter, with significant growth in fiscal 2025 and beyond.”
Given memory is where AI accelerators are competing in the near-term, Micron’s product road map is one to watch closely to see if the company can maintain market share and/or has the ability to expand.
China is an inherent risk to Micron with decently large exposure and uncertainty regarding the CAC’s decision in March 2023 to prevent critical information infrastructure companies from buying Micron’s chips remains uncertain. In 2023, China accounted for ~$2.18 billion in revenue, or 14% of total revenue, up from 10.8% in 2022 and 8.9% in 2021. Micron said that it believes “approximately half of that China-headquartered customer revenue, which equates to a low-double-digit percentage of our worldwide revenue, is at risk of being impacted.”
Micron Technical Analysis
By Knox Ridley
When analyzing a single stock, I always prefer to look at it within the context of the larger sector. The below chart shows the Philadelphia Stock Exchange Index (PHLX), which is composed of the 30 largest semiconductor companies in the U.S. It provides a decent snapshot into the health of the overall sector.
It currently shows the uptrend off the October 2022 low hitting significant resistance. The 4216 region has been our long-term target for this move higher, which happens to coincide with a 45-degree angle off the COVID low in red. This angle tends to be the dividing line between big trends. More times than not, we see a strong reversal around this angle.
From a cycle perspective, the above chart shows 5 cycles that have had an effect on PHLX. Note how they are clustering underneath price right now. The last time these cycles clustered together was at the 2021 high. More times than not, we tend to see a reversal into these cycles. The fact that both time and price are coming together for the PHLX signals caution here until we see how price reacts at this region.
Many semis have put in a larger top, while some suggest that we could see one more swing high into 2024, Micron being one of them. I do believe that the evidence supports MU putting in a notable top soon – likely in late Jan to late Feb. Once we reverse from this high, how we drop will determine how we buy.
The below chart has these two scenarios mapped out.
The blue count suggests that the drop will be a steep 5 wave pattern. If this happens, we will be targeting between $52 – $44 for the bulk of our buying.
The green count suggests that the drop will be a messy/three wave move. If this happens, we will buy between $65 – $55
The above two scenarios are what bets fit the larger pattern playing out. However, with a move as explosive as AI, we have to account for something more bullish that could alter the price pattern. This would be a vertical breakout above $99-$100. If this happens, you would see buy into that move with a stop just under the $96 region.
If we zoom into MU from the 2021 top, we can get more perspective on what is likely transpiring. It appears like we have an incomplete correction that started at the 2021 top. What gives this away is the uptrend we have seen in 2023. It is a messy/three wave move, which fits best as a correction in a larger downtrend. Until we see a large vertical move over $99-$100, the price action best fits with the blue and green counts I just laid out.
What is concerning in the momentum indicator on the bottom of the chart is in the same position that marked notable swings lower, including the 2021 top. This is happening while MU’s chart has room for, maybe, one more swing higher into the $90 or $92 region. If the next move extends, I’ll be looking for the $94.75 or $99 region for a top to unfold. If we instead break below $74, then I’d consider the top being in, and we will start analyzing the trend to determine where we buy more MU.
Conclusion
Fiscal Q1 offered more evidence that Micron’s recovery is underway, as HBM3e emerges as a significant enabler of generative AI applications and a significant growth driver over the next few years. Micron reported its first YoY increase for quarterly revenue in Q1, breaking a five quarter string of declines. GAAP gross margin is expected to see substantial improvement in fiscal Q2 to the low double-digits, while GAAP EPS is expected to shift back to positive by fiscal Q3. Operating cash flow has improved significantly, with positive free cash flow expected by the end of the fiscal year. Overall, Q1’s beat and raise are setting the stones for a strong fiscal 2024.
We’ve taken our time with this stock and created a game plan, which is detailed for you above. This game plan helps us to buy as low as reasonably possible without missing out on AI-driven price action. Pro Members will have ongoing earnings coverage on Micron and in-depth thematic coverage from our analyst team on Micron and other memory stocks. Advanced Members will receive a real-time trade alert when we layer-in and will receive weekly coverage of Micron in Knox’s webinars. To learn more about Advanced, click here.
I/O Fund Analyst Damien Robbins contributed to this analysis
AEHR reported a 20% cut at the midpoint to full year revenue guidance from $100M to $75M to $85M, — a near-perfect mirror of ON Semi’s 20% cut in its SiC revenue forecast from $1B to $800M.
We thought the impact could have been $16M given ON is 80% of AEHR’s revenue. Interesting enough, AEHR stated in the call the miss not from ON but rather from “other customer forecasts.”
We were cautiously optimistic that ON would not become a readthrough for AEHR, yet what has materialized is that the EV market is weaker than expected, and ON is one of many that will either report this or already has reported this. Therefore, it is irrelevant if ON created the miss as weakness in the EV market resulted in a similar miss.
This was the part in the Q&A where it was distinguished the miss did not only come from ON Semi or even primarily from ON Semi:
“Jed Dorsheimer
[…]it sounds like there was a material change in the last 30 days in terms of demand and visibility to your business. Is that correct? And I just asked that because your largest customer did flag something last quarter, but I want to separate what happened in the past versus what's occurred in the last 30 days. Thanks.
Gayn Erickson
All right. Well, one thing I'm going to start giving people heads up on is, I'm trying to be more thoughtful about giving insight into our customers, to be fair. But I'm going to specifically answer what I think you're implying. Right after our largest customer talked about their change in their forecast, we not only did not hear a negative impact to us, candidly, it flipped around and for a period of time, it was actually an uptick, right, which was a little hard to imagine and explain, but had to do with the wafer pack and the shift to new customers and some other things.
Literally in the last seven days, they have reconciled their plans, etcetera. And we've tried to thoughtfully reflect that in the latest one. But again, I actually said in my prepared remarks, their revenue to us is pretty close to what we were expecting when this all came out. So if you want to say the bulk of the $15 million to $25 million decrease was not from them, that was actually from other customer forecasts that have changed over the last, like three to four weeks candidly.”
This matched what was said in the press release:
“In the last sixty days, we have seen how the slowing of the growth rate of the electric vehicle market has had a negative impact on the timing of several current and new customer orders and capacity increases for silicon carbide devices used in them. […] The net of this is that we now expect a delay in the timing of new orders from current and new customers that will most likely impact this fiscal year’s revenue.”
In addition, Aehr says it believes it has a “large opportunity” with one of the market leaders in SiC, and while Aehr feels “confident they will move forward with our FOX-XP multi-wafer solution for their high-volume needs, [the] timing is taking longer than anticipated. We remain confident that we will receive initial purchase orders from them in fiscal 2024; however, it is not clear whether they will have the infrastructure ready to take shipments from us within our fiscal year that ends on May 31st.”
Revenue & EPS:
Revenue growth remained strong in Q2 as margins improved, though backlog and bookings decreased significantly QoQ. AEHR is lumpy with management announcing new orders intra-quarter, so we’ve seen these lows before on backlog and bookings, but it’s not exactly ideal.
Revenue increased 44.6% YoY to $21.43M, beating estimates by 2.6%
GAAP EPS of $0.20 increased 53.8% YoY
Non-GAAP EPS of $0.23 increased 43.8% YoY, beating estimates of $0.19 by 21.1%
Given the fiscal year guide was lowered by $20M at the midpoint, analyst revisions will likely look like this:
Q1 FY2024: $20.62M reported
Q2 FY2024: $21.43M reported
Q3 FY2024: $15M to 16M down from $20.89M consensus
Q4 FY2024: $22M to $23M down from $26.91M consensus
I’m taking this from a comment in the call that 40% of what’s left will be in Q3 and 60% will be in Q4. Per the CEO: “Q4 will certainly be bigger than Q3. Maybe but not majority, maybe at 60, 40 spreads or something like that.”
Margins:
Gross margin of 51.1% increased ~270 bp QoQ, but decreased ~230 bp YoY. Per the CFO, this was due to a high inventory reserve, up 24% from last year. This is a concern, should inventory continue to increase, it can weigh on pricing.
Operating margin of 25.5% increased ~550 bp QoQ and ~200 bp YoY. The QoQ increase in operating margin was primarily driven by a (19.7%) reduction in R&D expenditures relative to Q1.
Net margin of 28.4% increased ~570 bp QoQ and ~330 bp YoY, reaching the highest level since 28.6% in fiscal Q4 2022.
Cash & Debt:
Cash and equivalents of $50.5M, a marginal (1%) QoQ decline from $51M in Q1.
Debt remained at zero.
Operating cash flow is implied to be ($0.54M) in Q2, with Aehr reporting $3.9M in operating cash flow in Q1 and $3.36M in operating cash flow in the first six months.
Key Metrics:
Bookings were $2.2M in Q2, a QoQ decline of (88%). This represented the lowest quarterly bookings in more than eight quarters.
Backlog was $3.0M in Q2, a QoQ decline of (86%) and YoY decline of (81%). This represents the lowest backlog in more than eight quarters.
Earnings Call:
ON Semi Won’t Be the Top Customer Next Year:
The boldest comment on the call was not regarding the miss, but rather regarding the fact their number one customer will not likely be their number one customer in FY2025. This either means AEHR sees EVs being so weak that ON will reduce its SiC forecast again, or that other industries will step up and place new orders that exceed ON’s roughly $80M. It’s likely a mix of both scenarios. Here is what was said:
“Christian Schwab:
[…] So given their public comments, I guess, let's start with that, how would you anticipate that customer materiality in fiscal year 2025?”
“Gayn Erickson:
I mean, we — I believe that they will still be material. I don't know that they will be the dominant customer [Technical Difficulty] they'd be or they not be. My guess is, they will not even be the largest, as some of the other customers are kicking in with their ramps. One thing we've tried to look at is, how fast is the market growing itself. I mean, silicon carbide is growing, let's say 40% a year topline revenue. Can we grow faster than that? I think there is examples where we can, but I think it would be more realistic to think that we grow alongside the market itself. But as we displace potential package part burn-in, et cetera, there is opportunities […] But we do think that they will still be a significant customer for us next year. We believe that they — and are consistent with what they have been telling people their growth plans are. But we think that there'll be other customers, most likely there'll be bigger than them next year.”
My translation: I would rather wait to see what this transition looks like than speculate on how it will play out when ON Semi is no longer the top customer. I believe ON Semi will have to come down in revenue from where we currently are otherwise next year’s fiscal year would be higher as a customer that rivals ON’s current revenue contribution would have a larger impact. It comes down to the backlog being very slim and also bookings. We track key metrics for a reason, and it’s not prudent to ignore them.
Auto Inventory Levels:
Auto inventory levels for December are high. In the United States, the total supply of unsold new vehicles was up 57% from the same time a year ago. Inventory is 17 days higher than it was a year ago.
According to SP Global, Ford has cut its EV production for F-150 Lightening in half and Chevrolet EV inventory is up 19% QoQ.
SP Global is also reporting that dealer inventory across all vehicles is up 60% year-over-year to 2.3 million units.
Conclusion:
The main negative for Aehr’s fiscal Q2 stems from the reduced FY revenue outlook, marking a significant reset in growth expectation from the high 50% range to the high teens to 30% range. This was certainly a negative surprise but what’s most concerning is the low bookings and backlog as we don’t have any history of owning AEHR with key metrics that are not supporting the $38 million in orders the company has to procure by the end of the fiscal year in May. AEHR is capable of getting orders in quickly but we prefer to see evidence.
Margin improvement was a positive, with gross margins recovering above the 50% range, while reduced R&D expenses aided leverage down the line. There is no argument that AEHR has an extraordinary ability to operate efficiently for its size, and this is one of the reasons we monitor the company closely.
It’s likely we close the position and reopen again in six months. Today, AEHR is a silicon carbide story relying squarely on the EV market. Unfortunately, with what we know today, the EV market suppliers are weaker than expected, and AEHR commentary is hinting it could continue.
We prefer to wait for evidence of a recovery rather than speculate on when this will happen. We will monitor AEHR closely for when the company secures orders from other large customers and industries. In other words, we will wait for a breakout, which is most likely to happen in the early part of AEHR’s next fiscal year.
Ultimately, we want to focus on putting stronger horses in the stable over the next six months. We suspect AEHR will break the $18 level Knox outlined and we will respect this stop and revisit later.
Silicon Carbide is powering a revolution in electric vehicles and AEHR is at the forefront. The company is a beneficiary of the switch from Silicon-insulated gate bipolar transistor (Si-IGBT) to wide-bandgap Silicon Carbide MOSFETs for electric vehicles components. The result in switching to Silicon Carbide (SiC) is that charging is quicker and the range of miles for electric vehicles increases with SiC.
AEHR has seen a ramp in demand over the past three years because silicon carbide is replacing silicon in electric vehicles. Tesla was the first to adopt silicon carbide for the 2018 Model 3 by working with ST Microelectronics to add SiC MOSFETs to an inverter design. The result was a more compact, lighter inverter at 4.8Kg compared to Si IGBT inverters that weigh 2-3X more (8kg to 12kg). SiC inverters offer 97% efficiency, resulting in more range, and this is achieved without the need to increase battery capacity.
During Q1 earnings, management reiterated the FY 2024 guidance of at least $100 million, representing over 50% year-over-year growth.
Since that earnings report, AEHR’s largest customer, ON Semi, provided a guide on SiC for 2023 that is 20% lower than the original guidance. As a reminder, ON is about 80% of AEHR’s revenue, and we’ve pointed out in nearly every write-up on the stock that customer concentration is the primary risk. Quick math indicates that if ON misses by 20% on SiC, then AEHR would miss by about $16M if we figure $80M of AEHR’s revenue is from ON Semi.
The risk that is involved with such high customer concentration is what the market is reacting to, but ON Semi’s call was not as bad as it sounds.
Here is what ON’s management team stated:
“We are taking a very cautious approach as we are starting to see pockets of softness with Tier 1 customers in Europe working through their inventory and increasing risk to automotive demand due to high interest rates.”
And later stated, “[…] for the full year, a single automotive OEM's recent reduction in demand will impact our $1 billion target, and we now expect to ship more than $800 million of silicon carbide in 2023, 4x last year's revenue. In '24, we expect the growth of our silicon carbide business to double the market growth.”
Management stated they plan to “double the market growth” in 2024, and more details were discussed further in the call:
Gary Mobley
Hassane, I want to pin you down on your market forecast for silicon carbide for next year to really get an insight into what your expectations are. I know you cited in your footnotes of your presentation today, a lot of Yole forecast, and Yole is forecasting roughly 43% growth in silicon carbide for next year. So are you expecting to grow your silicon carbide revenue 80% next year? Is that the proper read here?
Hassane El-Khoury
Well, it depends on what the Yole is. But yes, we are expecting a 2x market.
The CEO stated this again during the call:
“Hassane El-Khoury
Yes. Silicon carbide is — what we are looking at is silicon carbide in '24, basically growing about 2x the market. So it's still on track to the target that we've put out in Analyst Day of growing 2x the market, that we still have that visibility in '24.”
From this point forward, the conversations were positive, including SiC having a high utilization rate that is higher than ON’s corporate average, and that SiC will remain supply constrained in 2024, which is a positive read through for a company like AEHR that supplies ON Semi with wafers and wafer equipment.
Here is what was said about the utilization rate for SiC versus silicon:
“Quinn Bolton:
Got it. And then on the silicon carbide business, you talked about overall utilization rates being managed down to 68% for the next few quarters to manage inventory. I assume, given the outlook for EVs still growing in the fourth quarter into next year, that the silicon carbide is probably immune from some of those lower utilization rates, but wanted to clarify that? And if utilization remains high in silicon carbide, could you actually see a scenario where silicon carbide moves above corporate average in 2024?
Thad Trent:
Yes, that's a good question. So the utilization for silicon carbide in Q3 was up, where silicon was down, and that pulled the total up. As we look forward, we don't see the silicon carbide utilization decreasing. We will be bringing on additional capacity next year to support '25 and beyond. But I don't expect that utilization to decline. I think it's the silicon that will actually decline, that gets us down into that 65% to — mid-60% to high-60% range.”
What the Market Wants to See Tomorrow
Given that ON Semi forecast a fairly weak Q4 for SiC with the $200 million miss, the market will want to see AEHR maintain its fiscal year guide. The selloff is anticipating that ON’s miss will flow through to AEHR in the upcoming report. That’s the roughly $16 million mentioned above.
ON Semiconductor’s Q4 guide missed analysts' expectations by 9%, but to reiterate SiC missed by 20%. This miss puts increased pressure on AEHR’s management to deliver a confident outlook for the remainder of the year.
Below, is an interesting comment from ON’s management team during the earnings call, which if I’m interpreting it correctly, would indicate that the miss on SiC may not affect ON’s strategic plans to build wafer inventory. If so, AEHR has a chance of being unscathed (I’m being ever the optimist here, but there is some indication this could play out, per the comment below):
“Hassane El-Khoury
Sure. So it is all — I guess all of our '24 by now. For the '24, we have visibility on exactly what program, what voltage, what volume and what mix we need. As far as the inventory, Thad talks about ramping strategic inventory for silicon carbide, we stage inventory primarily in, I would say, in two spots. One is blank wafers or substrate, wafer substrate, which is fully fungible across any customer, any platform with any volume. And as we get closer, we stage inventory at Epi, which is when we, I guess, partitioned with the voltage levels of the product.
So this is where we maintain inventory to give us full flexibility should the shift change. Because we've always said, when we would have one or two platforms at a customer, if one vehicle sells better than the other, the customer would want to shift while still using onsemi. So we give that flexibility to be able to shift on between platforms at a similar OEM or between OEMs. So the best place to keep that inventory is blank wafers and/or Epi.”
Ideally, we hear updates on new customers. On Semiconductor constituted 79% of FY 2023 revenue, down from 82% of FY2022. Management said in the last earnings call that the last two new customers did not need their wafers tested on the company’s system before moving forward compared to the requirement for the early customers.
Bookings will be closely watched as last quarter was weak (more below). Backlog growth is also lumpy but will be looked at closely.
Revenue and EPS
The company’s Q1 FY2024 revenue grew by 93% YoY to $20.6 million and beat consensus by 7.1%. Management went out of their way to remind investors that this is the strongest first quarter they’ve ever reported and that first quarter is typically seasonally weak. They also highlighted the record shipments of FOX WaferPak Contactors, a key component for testing wafers and also provides recurring revenue for the company.
Gayn Erickson said in the earnings call, “As we've noted before, our proprietary WaferPak Contactors are needed with our FOX wafer level test and burn-in systems to contact with the individual die on the wafer and are designed specifically for a given device. As our customers win new designs from their customers, Aehr eventually secures orders for new WaferPaks to fulfill these new wins. With each new design, our customers will need enough new WaferPaks to meet the volume production capacity need for those new devices.”As our customers win new designs from their customers, Aehr eventually secures orders for new WaferPaks to fulfill these new wins. With each new design, our customers will need enough new WaferPaks to meet the volume production capacity need for those new devices.”
The company is expected to report $20.89 million in the November quarter, representing YoY growth of 41%.
Q1 GAAP EPS was $0.16 and is up from $0.02 in the year ago quarter. Adjusted EPS grew by 260% YoY to $0.18 and beat consensus by 12.5%. Overall, earnings are expected to nearly double in FY2024, per management guidance of “GAAP net income of at least $28 million, representing earnings growth of greater than 90% year over year.”
Margins
Q1 FY2024 gross margin came at 48.4% compared to 42% in the same period last year and 51.5% in the previous quarter. The higher revenue percentage of WaferPaks helped to improve the margins, and on the other hand, the WaferPak automatic aligners that are built externally led to a decrease in the margins. When an analyst asked the CFO in the earnings call if the gross margin would return to 50%, he affirmed this by saying, “So we're still targeting 50% above the margin for the year, and that's what we're looking at.”
The operating margin was 20% compared to 4.3% in the same period last year and 25.3% in the previous quarter. The operating margin was also softer than the previous quarters as operating expenses increased 45.8% YoY to $5.9 million due to “increased headcount-related expenses to support our worldwide sales and marketing efforts and our R&D programs.”
Net income grew by 694% YoY to $4.7 million (due to small numbers). The net margin was 22.7% compared to 5.5% in the same period last year and 27.4% in the previous quarter.
Cash Flow and Balance Sheet
The operating cash flow was $3.9 million, down (28.6%) year-over-year. Cash flows have been lumpy in the past. The free cash flow was $3.6 million compared to $5.4 million in the same period last year. Cash and investments on the balance sheet were $51 million, up from $36.2 million in the year ago quarter and up 6.5% sequentially, with no debt.
Key Metrics
Q1 Inventory is higher than usual at $31.56 million, up $7.6 million QoQ. The management has stated their plan is to increase inventory to meet upcoming demand.
Bookings in Q1 were $18.4 million and declined (-3.7%) year-over-year. Bookings were up from $15.2 million the previous quarter. These are lumpy but can cause the stock to move quite a bit during a good quarter.
Q1 backlog of $22.3 million was up 14.4% year-over-year yet was down sequentially. The effective backlog of $24 million could have been better, given management had stated they received $15 million in effective backlog for Q1 during the Q4 earnings call in July.
Conclusion
Per Knox’s most recent 25-page Positions Report: AEHR’s technical pattern has always been a mess. If its earnings report is accepted by the markets, we should hold $20.45 and turn higher. If this happens, then the drop we just went through from its high was only a 3 wave pattern, and supports us pushing to new highs in a final swing higher. This is the green count in the chart, and once we break above $39, the odds will favor this scenario.
However, if we drop below $20.45, then we have a clean 5 waves down from the high, which will be concerning. If what follows is a 3 wave bounce, then we will be setting up for a bigger drop. If this happens, we will likely look to de-risk our Aehr position.
Technical Analysis by Portfolio Manager Knox Ridley, who holds a 1-hour webinar on Thursdays for Advanced Members.
Note: We are taking a break from Netflix in the Essentials plan for now, and instead initiated Bitcoin as the setup looks stronger at the moment. You can read this analysis here.analysis here.
Microsoft
Microsoft is a very mature pattern. Our long-term target is $378, which it is struggling to push through. This target completes the large uptrend that started in 2009, so a sizable pullback is expected when the current rally ends – sometime between now and late March/early April, is our best guess.
Regarding MSFT, if still needs to complete that final 5th wave swing higher. This could target between $400 – $415. Please understand the risk involved with this great company and where we are in the trend. We plan to look lower into 2024 for better entries, and plan to hedge the remainder of our position.
Nvidia
Though it looks likely that the October 2022 low was a major low, we should see a sizable pullback into 2024/2025. However, before that, Nvidia looks like it needs to push towards the $545 – $600 level to complete the large 5 wave pattern off the 2022 low. The risk here is that the red alternative top in the chart below does satisfy the minimum requirements for a 5th wave. Below $480 and risk becomes quite elevated that this is playing out.
Advanced Signals Members receive real-time trade alerts for our entries and in-depth technical analysis from the Portfolio Manager, Knox Ridley. Learn more here.here.