Below, we look at the PC-related miss and what to expect from AMD in the data center.
The data center is critical right now because this is what can help AMD reclaim its stock price. I wouldn’t dismiss the possibility that this year becomes AMD’s best year in the data center as the company is perfectly positioned to make a substantial increase in market share over Intel. By best year I am not referring to growth percentage but rather judging by impact of total revenue and size of this segment. The move from AMD owning “mid-20%” of the CPU data center to owning 40% to 50% of the market is the move we want to capture — and this is entirely possible due to Intel’s most recent stumble.
Source: Tom’s HardwareTom’s Hardware
Pictured Above: AMD has grown from 2% market share to “mid 20-percent market share”
It’s important to note that AMD has grown over 6% of its market share in the last two quarters. Per our recent Q2 coverage: As an analyst pointed out, AMD appears to have gained 6% market share, which is “the highest share gain in the data center business that [AMD] has reported even going back to 2005.”
We have been quite thrilled to see the team at AMD led by Lisa Su and Forrest Norrod overtake Intel at times. However, what happened last quarter with Intel’s stumble is an exponentially greater mistake than the last stumble that we prepared for in March of 2020, which later materialized in July of 2020.
You may have seen tech commentary poking fun of Meta’s most recent keynote on social media and in newsletters. If you missed it, the recap is that Mark Zuckerberg demonstrated adding legs for Metaverse avatars and quickly became a target as the progress doesn’t quite reflect the company’s large level of investment.
Most investors look at Facebook’s cash and think “this will make a great stock,” which is first level thinking — yes, the FCF margin is impressive. Some are taking this further and scrutinizing “where is the cash being spent and will Meta succeed?”
However, the most important takeaway to Meta’s insatiable appetite to find the next big thing in media is that AMD stands to greatly benefit whether Meta succeeds or not.
Here is data presented on the forum from last year that shows Meta’s increased capex following the Q3 2021 earnings call.
We don’t get enough transparency to separate real estate costs from servers and data centers in the capex number but we do know that Meta chose AMD to build its data centers – which was an earth shattering announcement for AMD investors. Amazon did at one point comment about its split and said 40% of its capex is spent on data centers.
This is precisely why AMD leap frogging Intel is critical for investors to not forget or lose sight of in a consumer led selloff. Meta capex spending with AMD can grow to exceed AMD's PC revenue in the next 2-3 years alone, and of course, AMD is partnered with many hyperscalers beyond Meta.
The decision to choose AMD at Meta’s primary supplier was very likely due to AMD’s historic product lead over Intel, which we have covered in this webinar and this write-up. Point being, not only is AMD in the lead on product but AMD is in the lead at exactly the right time. Meta helps to quantify the impact.
In regard to Nvidia, one thing to remember, is that Nvidia’s Omniverse creeps onto Meta’s territory for building 3D Worlds. I imagine Meta’s ultimate goal is to attract creators and developers to build their 3D apps with Meta and this is also Nvidia’s Omniverse’s goal. AMD does not compete here and so it’s an obvious choice for Meta to partner with AMD on both CPUs and GPUs.
Brief Note on Data Center Numbers:
There is additional evidence of the impact of Intel’s previous stumble as AMD’s data center segment grew 83% year-over-year in Q2 compared to Intel’s (16%) decline. Intel has not caught up to the AMD’s products released two years ago and now AMD is being even more aggressive as the 5-nanometer line up is being released in Q4 which includes Zen-4 architecture plus the Zen-5 architecture in 2024. We covered the upcoming product road map here.
The company also stated that the Zen-3 Milan Series is still outstripping supply with visibility six quarters out, implying for full year 2023. As a reminder, Zen-2 was Lisa Su’s comeback and Zen-3 is responsible for the current move in data center market share.
However, as an AMD investor, I am actually less concerned with shiny numbers like 83% growth and I am more concerned with the data center segment becoming so large in total revenue that more cyclical segments like PCs have less impact on the stock.
Data center revenue is a $6 billion segment with Q2 data center revenue at $1.5 billion and for Q3 is $1.6 billion. AMD did not break out the data center number before but judging by the 83% in Q2 and the 45% in Q3, the company looks to be tracking in the ballpark of Meta’s capex number of 66% (I find the average growth for AMD to be roughly close to Meta’s capex if we assume Q1 was in line with Q2 and that Q4 will be roughly in line with Q3 to be interesting although there isn’t enough information about Meta’s orders to verify the correlation).
If Meta reports 30% growth in capex next year, the data center segment can easily become $7.8 billion next year. However, what this overlooks is the impact of AMD taking more market share from Intel, which will likely be seen when AMD attracts higher budgets from more hyperscalers. If AMD takes another 10% to 20% of market share over the next 18 months, we could see roughly a $9 billion to $10 billion segment, which would be 60% growth over 1.5 years (or 40% average annual growth). This is entirely possible as AMD is taking 12% market share this year, and again, I believe AMD eating into Intel’s market will actually accelerate more this year than it did last year. I am using the word “year” loosely to say Q4 2022 through the end of fiscal 2023, so about 6 quarters.
I do want to point out that Arm is starting to creep into X86 server territory as evidenced by the chart above, however, I believe this is a bigger problem for Intel and not as much for AMD who has clearly not been impacted by Arm’s growth.
More on Big Tech Capex
Royston wrote about Big Tech Capex on the forum and some of this is going out as our free newsletter this week. Here is what he wrote:
Our team has been working on looking closer at AMD and here is some research from Big Tech Capex, which was a leading indicator going into 2022. Our goal is to see if there's any sign of slowing data center and server investments from the Big 4.
Monitoring the capital expenditures of big tech companies helps to understand the demand for data centers and artificial intelligence. While not all Capex goes to data centers and AI, it remains a significant investment in recent years.
Meta’s capital expenditures in Q2, including principal payments on finance leases were $7.75 billion, up 64% YoY. The company’s CFO, Dave Wehner, said in the Q2 earnings call, “Capital expenditures, including principal payments on finance leases, were $7.7 billion, driven by investments in servers, data centers and network infrastructure. The big step-up in CapEx, both year-over-year and sequentially related to server spend, including for our AI infrastructure.”driven by investments in servers, data centers and network infrastructure. The big step-up in CapEx, both year-over-year and sequentially related to server spend, including for our AI infrastructure.”
Sheryl Sandberg said, “Third, on AI and machine learning. I want to emphasize Mark's point that this is a really important part of how we improve our ads ranking and measurement capabilities. AI-driven products like advantage detailed targeting and advantaged look-alikes help to increase the audience for an ad campaign if it's likely to improve performance. AI is also an important part of how we continue to grow video monetization.”AI and machine learning. I want to emphasize Mark's point that this is a really important part of how we improve our ads ranking and measurement capabilities. AI-driven products like advantage detailed targeting and advantaged look-alikes help to increase the audience for an ad campaign if it's likely to improve performance. AI is also an important part of how we continue to grow video monetization.”
The company expects 2022 capital expenditures, including principal payments on financial leases, to be $32 billion at the mid-point of the guidance, representing a 66% YoY growth. Tracking the Capex in the first two quarters, Meta Platforms had spent $13.3 billion. It suggests it will be higher in 2H 2022, as when we deduct from the mid-point of the guidance, it comes to $18.7 billion.
The company is planning to reduce the hiring of engineers by 30% this year due to the economic slowdown according to an internal memo. The company is facing the heat due to privacy issues and slowing ad revenues. The memo suggests a five-fold increase in the requirement of GPUs to make its feed better to increase engagement using Artificial Intelligence as the company is trying to combat the competition from TikTok.
Meta also recently announced its plan to expand the Eagle Mountain data center project. It is Phase 3 expansion plan and brings the total investment in the project to over $1.5 billion.
Alphabet’s Q2 Capex grew by 24% YoY to $6.9 billion. Ruth Porat, CFO of Alphabet, said, “Turning to CapEx. The largest investments in the second quarter were in servers followed by data centers and office facilities.” were in servers followed by data centers and office facilities.” The company had invested $24.6 billion in Capex in the year 2021, up 11% YoY. The management expects Capex to rise in 2022. In the Q2 2022 earnings call, Ruth Porat said, “We continue to expect an increase in CapEx in 2022 versus last year. For the balance of 2022, the increase will be particularly reflected in investments in technical infrastructure globally with servers as the largest component.” Earlier this year, the company announced its plan to invest about $9.5 billion in data centers and offices in the U.S. for the year 2022. This is up from about $7 billion spent in 2021.
Similarly, Microsoft’s Capex including financial leases, grew by 19% YoY to $8.7 billion in the Q4 FY2022 quarter (i.e., Q2 CY2022). Amy Hood, CFO of Microsoft, said, “Maybe let me start by talking about Q4's capital spend. Obviously, the big driver of our growth this quarter was in data center spend, both new and newbuilds as well as adding capacity to existing data centers. We are seeing, obviously, good demand signal.” data center spend, both new and newbuilds as well as adding capacity to existing data centers. We are seeing, obviously, good demand signal.” The management expects a sequential decrease in the next quarter due to the normal variability in quarterly spend. In the CY 2021, Microsoft’s Capex including financial leases, grew by 33% YoY to $27.5 billion.
Amazon incurred capital expenditures, including equipment financial leases, of about $60 billion in 2021. About 40% of this is made up of technology infrastructure supporting AWS and worldwide stores business. The management expects Capex to increase over the last year with the increase in technology infrastructure.
Brian Olsavsky, senior VP and CFO, said in the Q2 2022 earnings call, “For full-year 2022, we do expect to spend slightly more on capital investments than last year, but the proportion of capital spending shifts among our businesses. We expect technology infrastructure spend to grow year-over-year, primarily to support the rapid growth in innovation we are seeing with AWS. We expect infrastructure to represent a bit more than half of our total capital investments in 2022.”
AMD’s PC-Miss
As you already know, AMD pre-announced earnings with a miss of $1.1 billion for revenue of $5.6 billion, down from previous guidance of $6.7 billion. This will represent growth of 29% compared to growth of 55% at the mid-point of the guidance that was previously expected.
The low margins would also imply that average sales price is coming down quite a bit in addition to number of units. Here is what IDC said: “Shortages over the last several years have aggressively driven product mix shifts towards the premium end. This, coupled with cost increases of components and logistics, drove ASPs up five quarters in a row to $910 in 1Q22, the highest since 2004. However, with demand slowing, promotions in full swing, and orders being cut, the ASP climb was reversed in 2Q22. Another quarter of ASP declines indicates a market in retreat."
Although AMD has a clear lead ahead of Intel in the data center (where it matters most for long-term investors) this is not necessarily the case in PCs. According to Tom’s Hardware, the AMD Zen 4 series can compete against Intel’s current Alder Lake. However, AMD’s Zen 4 Ryzen 7000 coming out this month will have its hands full competing with the anticipated Raptor Lake from Intel due out later this month. You can read about the benchmarks here but the main takeaway is that average sales prices comes under pressure due to a more competitive market in PCs and gaming. This is in addition to the (15%) to (19.5%) decline in shipments noted below. Hence, we are seeing a much larger decline in revenue (40% decline in AMD’s revenue) driven by both lower ASP in addition to lower shipments.
A side note to consider is that the lower shipments are not only driven by slowing consumer demand. Some of this is coming from enterprise budgets, which raises questions around how fast enterprise budgets are deteriorating if AMD had little foresight into this issue.
I also want to note that enterprise can weigh on AMD’s data center segment even if Big Tech is a tailwind. If enterprises, government or education sectors are cutting server budgets than some effects from this will be seen on a YoY basis.
It can be tough to draw hard and fast conclusions is because comments about Q4 and next fiscal year are not available until November 1st. Operating from this unknown, I have the following information to share:
The most important question about the PC miss from AMD is when will we see a bottom?
There are two perspectives on this, the first is that Q3 is the bottom and the second perspective is that Q2 2023 will be the bottom. I imagine Q4 will be supported by the holidays to some extent, yet I would plan on lumpiness to continue in the Client category.
I also want to point out that estimates often have errors and this is true even for the very best analysts that go on record about PC sales. For example, IDC shows Apple shipments at 10 million for Q3 for 40% YoY growth whereas Gartner shows Apple’s global shipments at 5.8 million — this is a nearly 100% discrepancy between two reputable analyst firms.
IDC is reporting (16%) growth for the PC market in Q3 whereas Gartner is reporting (19.5%). The difference is primarily in the Apple numbers. Regardless, the PC market contracted 17% or 18% at the midpoint.
There are also institutional analysts who have numbers that are likely far off the mark: “September notebook shipments were up 12% month-over-month, well below Citi's expectation of up 23% due to continued inventory digestion and demand deterioration, Citi analyst Christopher Danely tells investors in a research note. As such, Q3 notebook shipments were up 3% quarter-over-quarter, well below the analyst's recently lowered forecast of up 7% quarter-over-quarter.”
This doesn’t matter for this quarter as AMD has provided the guidance but it is important to take into consideration for any upcoming forecasts.
Analysts who think Q3 will be the Bottom – Either PC Sales or Price:
Please note: the reason we look to institutional analysts particularly for a PC-related revenue miss is they have the ability to do channel checks and have well established contacts in the semiconductor industry given it’s a multi-decade market.
Jefferies analyst Mark Lipacis noted that AMD negatively preannounced Q3 revenues 17% below consensus, which is "entirely attributed" to client PC revenues being down 53% quarter-over-quarter and 40% year-over-year. Revenues for other segments, including datacenter, were in-line and he thinks Q3 revenues will "be near if not at the bottom," Lipacis tells investors. Given his belief that the stock's 39% underperformance versus the S&P 500 since November 30 discounts the bad news, he advises investors to be buyers of AMD and keeps a Buy rating with a $135 price target on the shares.
BofA analyst Vivek Arya lowered the firm's price target on AMD (AMD) to $90 from $100 and keeps a Buy rating on the shares after the company warned Q3 sales would be $5.6B, versus a prior $6.7B outlook. AMD did not update its Q4 outlook, but he expects trends to stay sluggish and models sales to decline further quarter-over-quarter on Client segment weakness, Arya said. He believes AMD's warning will have the most negative read-across for "PC peer" Intel (INTC), but also somewhat for Nvidia (NVDA) due to its exposure to consumer graphics and related memory and data center peers. He expects most semi cuts to be reflected when companies report Q3 results, which "could conceptually help create a trough in semi stocks," assuming the macro environment doesn't get worse, Arya added.
Bullet Points for a Bottom in/around Q2 2023:
Stifel analyst Ruben Roy lowered the firm's price target on AMD to $100 from $122 and keeps a Buy rating on the shares after the company announced preliminary Q3 revenue that was much lower-than-expected, driven by much lower-than-expected Client segment sales. He now forecasts Q3 non-GAAP EPS of 69c, versus a previous estimate of $1.06, but expects new product ramps and continued market share gains to drive a re-acceleration of revenue growth in the second half 2023, Roy said.
IDC stated that the PC and tablet market is forecast to decline (2.6%) before returning to growth in 2024. However, this Forbes editorial from a PC-veteran stated: “in my discussions with ODMs, the makers of PCs primarily in Taiwan and China, they suggested that for the enterprise market, we could see growth by as early as Q3 in 2023. In addition, they tell me that while consumer demand most likely will not rebound until 2024, they say they are now seeing increased orders for enterprise-based PC and laptops for delivery as early as Q3 of 2024.
PC vendors are getting comments from enterprise customers who held off buying large quantities of PCs and laptops during the last two years, and many of their PCs designated for IT use are well beyond their turnover date. With that in mind, some vendors are increasing their orders for IT-targeted PCs and laptops for delivery in late 2023.”
Northland analyst Gus Richard lowered the firm's price target on AMD to $80 from $105 and keeps an Outperform rating on the shares. Richard cut his 2023 estimates for AMD client revenue and data center revenue by $3.8B and $1.0B, respectively, due to the rapidly deteriorating PC market and declining enterprise spending following last week's pre-announcement.
My note: you can see the potential effects from enterprise budgets in Gus Richard’s estimates and this does include government and education budgets.
Note on the Other Segments:
We have 2024 tagged as the potential breakout year for automotive. If it happens sooner, we won’t complain. But, according to the product road maps that I’m tracking, 2024-2025 seems to be the sweet spot. We will table this for now but continue discussing in ongoing earnings coverage.
Regarding gaming, please note that AMD does not have exposure to crypto mining like Nvidia. On the flip side, Nvidia does not supply the CPU PC market like AMD. However, AMD does compete in the client-side GPU market which is reflected in the Client segment.
Conclusion:
Owning semis has been tough last few months, however, our research indicates Big Tech capex may be one of the strongest growth areas next year (again). We have to consider that outside of cybersecurity, there are going to be very few growth markets in tech in 2023, and of the growth markets we are tracking, very few will be in the double digits. What I’m referring to is an increase in budgets and/or spending in the areas that the tech participates in.
I like how Timothy Morgan from Next Platform worded the hybrid business model in his most recent write-up on AMD, and he also touches on Nvidia here:
“The numbers from both AMD and Nvidia demonstrate all too well how important it is to sell different kinds of compute engines across a wide variety of markets. While it is true that with such a hybrid approach there always seems to be something going wrong that limits profit and revenue growth, that hybrid vigor is nonetheless what helps a company make it through all of its hard times. IT suppliers who forget this – IBM and Hewlett Packard come immediately to mind – pare themselves down to a very tight market that makes them necessarily less resilient. And sometimes less relevant. Microsoft and Dell, and to a lesser extent Cisco Systems, seem to understand this.”
Our portfolio is centered on companies taking more market share from a competitor (AMD/Intel) or taking nearly all market share in one of the most expensive components of the data center (GPUs/NVDA). It’s not out of the question that the data center growth rates for AMD and NVDA outperform next year to help absorb the consumer-related weakness. Our best leading indicator for these stocks has been and will continue to be Big Tech capex.
Big Tech capex is important as it’s the one catalyst that can raise revenue estimates for next year, which subsequently raises bottom line estimates. Right now, the growth rate for AMD is quite abysmal at 3% for Q1 2023 so we want to see if we can get increased capex and then some revisions for H1 2023 (is the #1 thing I’m watching right now).
We will leave it to social media to stir up noise and distractions on the quality of Mark Zuckerberg’s avatar legs — and meanwhile, we will be busy digging up those capex numbers from Meta and other Big Tech companies. Last year, we got quite a bit of information regarding Big Tech capex from the October earnings reports, so you’ll be hearing from us if we get those updates in the next two weeks.