AMD is seeing weak price action for two reasons. First, the company is guiding for a decline QoQ from Q4 to Q1 across most segments, including the data center. Management also implied H1 of 2025 will be flat compared to H2 of 2024 in the data center. The market’s reaction is saying, “this isn’t exactly Nvidia
Second, management is dropping AI guidance from their commentary. If you’ve been following along, then you’re aware that AMD provides guidance around their expectations for AI-related GPU sales. For example, they stated this year GPUs would “exceed $5 billion” yet we started the year with commentary that GPUs would be in the range of $2 billion.
In response to AMD’s qualitative guidance, analysts would raise estimates, making it impossible for AMD to provide a surprise beat. It was a tough situation, and fairly rare to have analysts always doubling what a management had stated (on a forward basis). Therefore, I don’t disagree with what management is doing, which is creating immediate-term pain for a more equal playing field as they go ship the better architecture in the second half of 2025.
The Street thinks dropping AI guidance is hiding an issue, and communicating that there must be more weakness ahead. Almost-always, when a management team drops a key metric, it means it’s too weak to continue reporting this key metric. I believe it’s the opposite, that AMD is setting up the equal playing field before the MI350 and MI400 architecture arrive, as it’s been generally understood to be the moment when AMD can close the gap with Nvidia on some workloads.
The topic of “some workloads” is where the market will continue to get the AI story wrong, in my opinion. There is not one perfect chip, or AI system, for every single workload. It’s true that to train the trillion+ parameter models coming out of OpenAI that Nvidia’s NVL systems will be needed. However, it’s my contention that as some Big Tech workloads and enterprises seek lighter models at a reduced cost, that AMD’s new GPU architecture (coming in H2 2025) will be a strong choice.
Lastly, AMD is in pole position for edge computing, which they collectively refer to as mobile processors. Similar to data center GPUs, Nvidia will bring its sheer might and compute power as evidenced by Project Digits (the $3,000 supercomputer), yet you will you hear me continue to recognize AMD as an AI PC leader, and for lighter AI workloads at the edge (and those not looking to blow the bank), this is unlikely to change anytime soon as it’s what AMD does best.
Regarding the weak price action, overall, our firm has done quite well on this stock over the years and even recent quarters. It’s understandable if some investors are growing frustrated, yet this is not a stock we hold at a loss. We think we may be identifying a bottom, and this is key to understanding why we decide to continue to hold a stock, at times. Knox can fill you in more on Thursday on what he is seeing. If we think the stock has bottomed, then you’ll likely see us tough this one out as we are overweight Nvidia in both its stock ticker and its suppliers. We are strategically preparing for the days of “cheaper AI” which has been a thesis of ours for years; I’ve stated on Fox recently that Nvidia’s biggest threat is its gluttonous pricing power.
On the topic of custom silicon, AMD hinted its hat is in the ring. As is Nvidia. As is Broadcom, of course. As is Marvell, of course. As are a few startups. Due to the sensitive nature of IP, I foresee a scenario where each hyperscaler uses a different design company as it could expose their IP to go to one source for custom designs. We will track this as we go along.
Data Center Revenue falls short of estimates
AMD beat the revenue and EPS estimates. The company’s Q4 revenue grew by 24.1% YoY to $7.66 billion, beating estimates by 1.8%. The adjusted EPS grew by 41.6% YoY to $1.09, beating the estimates by a modest 0.4%. The revenue guidance for Q1 is $7.1 billion, representing YoY growth of 29.7% at the midpoint and beating the estimates by 1.6%.
The company’s revenue growth was led by record quarterly data center and client segment revenue. Data center revenue grew by 69% YoY and 9% QoQ to $3.86 billion. However, the revenue fell short of estimates of $4.14 billion. AI revenue exceeded $5 billion for the year 2024. Management expects strong AI and data center revenue growth in 2025. However, the earnings call lacked visibility on when AI revenue would ramp up, and guidance was missing in the earnings call.
The company’s launch of MI350 chips in mid-2025 should be a significant catalyst as the management believes that the MI350 chips have the largest generational increase in AI performance they have ever produced.
Revenue
The company’s Q4 revenue grew by 24.1% YoY to $7.66 billion, beating estimates by 1.8%. The company’s revenue growth was led by record quarterly data center and client segment revenue, partially offset by lower embedded revenue.
- Management revenue guidance for Q1 is $7.1 billion, representing YoY growth of 29.7% at the midpoint and beating the estimates by 1.6%. Management mentioned that the revenue will be down (-7%) sequentially due to seasonality.
- Analysts expect 28% growth in Q2 and 23.2% growth in Q3.
- The full-year revenue grew by 13.7% YoY to $25.8 billion.
- Looking further out, analysts expect revenue to grow 23.9% YoY in 2025 and 23.3% in 2026.
- Management expects double digit revenue and EPS growth driven by strong AI demand. The second half revenue is expected to higher than the first half due to the MI350 launch in mid-2025.
Segments
Data Center
Data Center revenue grew by 69% YoY and 9% QoQ to a record $3.86 billion. However, the data center revenue fell short of estimates of $4.14 billion and also decelerated from 122% growth in Q3. Management highlighted during the earnings call that 2024 was a significant year for the server business driven by the ramp-up of fifth-gen EYPC Turin and strong double-digit YoY growth in fourth-gen EPYC sales.
Data Center segment operating income was $1.2 billion or 30% of revenue compared to $666 million or 29% a year ago.
During the Q&A, there was a discussion that GPUs could be about $2 billion (an analyst’s words, not management – but they did not deny this number). Yet, the sticking point (as discussed) is that H1 will be flat compared to H2. Here was a response to his question as to whether the $2 billion will grow in H1: “I think you should assume that the first half of 2025 data center segment will be consistent with the second half of '24. And that's true for both businesses on the server side as well as the data center GPU side.”
Client Segment
Client segment revenue grew by 58% YoY and 23% QoQ to $2.31 billion, driven by strong demand for Ryzen desktop and mobile processors.
Client segment operating income was $446 million or 19% of revenue compared to operating income of $55 million or 4% of revenue a year ago, driven primarily by operating leverage.
Per management, they are not concerned that inventory is building with these knockout numbers. “We don't believe there is some substantial inventory build up. We actually think that what we're seeing is very strong adoption of our new products. So on the desktop side, we saw our highest sell-out in many years, as we went through the holiday season, launching our new gaming CPUs, frankly, they have been constrained in the market, and we've continued shipping very strongly through the month of January as we are catching up with some demand there.”
Gaming Segment
Gaming segment revenue was $563 million, down (-59%) YoY, primarily due to a decrease in semi customer revenue. However, was up 22% sequentially. Gaming segment operating income was $50 million or 9% of revenue compared to $224 million or 16% a year ago.
Looking forward, management believes channel inventories have now normalized and semi-custom sales are expected to return to more historical patterns in 2025.
Embedded Segment
Embedded segment revenue was $923 million down (-13%) YoY and flat sequentially due to the slower recover in the end market. Embedded segment operating income was $362 million or 39% of revenue compared to $461 million or 44% a year ago.
Margins
The company’s margins are improving with higher data center revenue mix. Going into 2025, management expects margins to be better in the second half of the year due to higher contribution from the data center revenue and it was clarified during the earnings call Q&A. The client business expansion will be a headwind in the first half due to the higher concentration of consumer business, which has margins below the corporate average.
Toshiya Hari (Analyst)
“That's great. And then as a quick follow-up, maybe one for Jean. So you're guiding gross margin to 54% in the first quarter. I'm curious what some of the major puts and takes are and those are the things that we should be cognizant of going into Q2 and more importantly, the second half. Given your data center commentary skewed more to the second half, I would expect margins to improve in the second half. But yes, if you can kind of run through the pluses and minuses, that would be really helpful. Thank you.
Jean Hu (CFO)
Yes. Thanks for the question. You are right. Our gross margin is primarily driven by our revenue mix, I think when you look at looking to 2025, Q1 guide, not only data center continued to grow significantly year-over-year. At the same time, client business is also growing year-over-year. So overall, the revenue mix is quite consistent with the Q4. So the gross margin guide is 54%. I think for the first half, if the revenue mix is at this level, we do feel the gross margin will be consistent with 54%. But going into second half, we do believe the data center is our fastest growth driver for the company and that will drive the gross margin to step up in second half.”
- Q4 gross margin was 51% compared to 47% in the same period last year. Adjusted gross margin improved 300 bps YoY to 54% due to favourable shift of higher mix of data center and client revenues, lower gaming revenue and partially offset by lower embedded revenue. Management guide for Q1 is 54% due to the similar revenue mix as Q4.
- Operating margin was 11% compared to 6% in the same period last year. Adjusted operating margin was 26% compared to 23% in the same period last year. Operating expenses increased 23% YoY to $2.1 billion as the company invested due to the expected strong growth in AI. Management expects adjusted operating margin to be 24% in the next quarter.
- Net margin was 6% compared to 11% in the same period last year. Adjusted net margin was 23% compared to 19% in the same period last year.
EPS
The adjusted EPS grew by 41.6% YoY to $1.09, beating the estimates by a modest 0.4%. The strong growth was driven by higher data center revenue.
- Analysts expect adjusted EPS to grow 52.2% YoY in Q1 and 55.7% in Q2
- Looking further out, analysts expect adjusted EPS to grow 41.5% YoY to $4.68 in 2025 and 46.3% YoY to $6.85 in 2026.
Cash Flows and Balance Sheet
The company’s cash flows are improving driven by higher profits.
- Q4 operating cash flow grew by 241% YoY to $1.3 billion or 17% of revenue compared to 6% in the same period last year driven by higher data center revenue.
- Q4 free cash flow grew by 351% YoY to a record $1.1 billion or 14% of revenue compared to 4% in the same period last year.
- Cash and short-term investments were $5.13 billion and debt of $1.72 billion compared to $4.54 billion and $1.72 billion at the end of Q3.
- The company repurchased shares worth $256 million in Q4.
Earnings Call:
Dropping AI Revenue Guidance:
AI revenue exceeded $5 billion for the year 2024. Despite the change in tone and dropping AI guidance, there was one exchange in particular where management reiterated it will grow “strong double digits” in 2025. The details can appear confusing, as on one hand, you have H1 2025 as flat, but on the other hand, you have the growth being “strong double digits.” For anyone who follows AMD closely, it’s assumed they are referring to a strong MI350 launch in what appears to be between May-July time frame (that’s my take on the timing, considering what management has described – typically it would have been August for AMD’s cadence).
This question cut to the chase of what was on everyone’s minds – which is how management is thinking about the impact of MI350s coming given a flat H1 2025 had been discussed.
“Stacy Rasgon
Got it. Thanks. And I guess for my follow-up, maybe to follow on there, do you think your exit rate on GPUs in '25 is higher than your exit rate in '24. Are you willing to commit to that?
Lisa Su
Absolutely. But yes, of course. It would be hard to grow strong double digits otherwise, right?”
As stated in the opening remarks, the launch of MI350 chips in mid-2025 should be a significant catalyst as the management believes the chips have the largest generational increase in AI performance they have ever produced. There was additional information provided:
“So as it relates to how data center — so the overall data center business will grow strong double digits certainly, both the server product line as well as the data center GPU product line will grow strong double digits. And from the shape of the revenue you would expect that the second half would be stronger than the first half, just given MI350 will be a catalyst for the data center GPU business. But overall, I think we are very pleased with the trajectory of the data center business in both 2024 and then going into full year 2025.”
Q1 Data Center Weakness:
Although Q4 beat on revenue, Q1 is expected to be soft with a 7% sequential decline. According to management, data center will be down just above the (-7%) total revenue sequential decline and in line with the corporate average, client and the embedded business to be down more than the total revenue decline, gaming to be down less than the total revenue decline.
The moment the stock went from being down 3% to being down 8% was when more color was provided that data center would be participating in this decline. Client is expected to be weak from Q4 to Q1, as device sales are always lower in Q1 for all consumer-device related companies (i.e., Apple, etc). Yet, seeing Gaming having stronger QoQ growth than data center is not ideal.
Here was the conversation that caused the stock to see steeper losses:
Q: Aaron Rakers (Analyst)
Yes. Thank you very much. And as a quick follow-up, just thinking about the guidance overall relative to that down 7% sequential I know you mentioned seasonality across the business segments. Are you assuming that you are down sequentially in data center in total in 1Q? And how do I frame that relative to seasonality? Thank you.
A: Lisa Su (CEO)
Yes, sure, Aaron. So let me give you some more color on the Q1 guide. So Q1 guide was down 7% sequentially, as Jean mentioned. And the way that breaks out in each of the segments assume that data center would be down just about that average, so the corporate average. We would expect the client business and the embedded business to be down more than that. Just given where seasonality is for those businesses. And then we would expect gaming business will be down a little less than that. And that's a little atypical from a seasonality standpoint, but we are coming-off of a year when there was a lot of let’s call it, inventory normalization. And now that inventory has normalized, we would expect that, that would be down a little bit less than the corporate average.”
MI350s Shipping Sooner than Expected
As with all underdogs, AMD has only a few chances to make its mark. We’ve been reporting for some time that this would be the CDNA-4 architecture with the MI350s (shipping in volume H2 2025), and if successful, then the CNDA-NEXT with the MI400s will be another opportunity (probably about a year later, H2 2026).
Here is what I’ve stated in the past:
“We will match that timeline and say we hope to see AMD be a leader in the market and in our portfolio by 2027-2028. CDNA 4 architecture is due out in 2025-2026, and is the most likely catalyst that I see today to narrow the product road map with Nvidia.”
This is what management stated about the upcoming architecture: “CDNA 4 will deliver the biggest generational leap in AI performance in our history, with a 35 times increase in AI compute performance compared to CDNA 3. The silicon has come up really well. We were running large-scale LLMs within 24 hours of receiving first silicon and validation work is progressing ahead of schedule.”
The silver lining to the earnings report is that it looks like investors will get a guide that includes some of the MI350s in the next earnings report as AMD is moving up their shipping expectations. Previously, this would have been about two quarters out. From Lisa Su: “Based on early silicon progress and the strong customer interest in the MI350 series, we now plan to sample lead customers this quarter and are on track to accelerate production shipments to mid-year.”
There were additional comments in the Q&A on this timing:
“And then the big news is on the MI350 series. So we had previously stated that we thought we would launch that in the second half of the year. And frankly, that bring-up has come up better than we expected, and there is very strong customer demand for that. So we are actually going to pull that production ramp into the middle of the year, which improves our relative competitiveness.”
By relative competitiveness, what AMD’s goal is put it into plain words, is that while Blackwell is greatly supply constrained due to the massive NVL systems, and already sold-out next year, AMD will aim to release a competitive architecture that serves the outsized demand for the B200s.
Quick Note on Custom Silicon:
There are many competitors on custom silicon, and AMD made it clear in the earnings call that they are one of them:
“And I just want to reiterate on the ASIC side, look, I think ASICs are a part of the solution, but there — I want to remind everyone, they are also a very strong part of the AMD sort of toolbox.
So we've done semi-custom solutions for a long-time. We are very involved in a number of ASIC discussions with our customers as well. And what they like to do is, they'd like to take our baseline IP and really innovate on top of that. And that's what I think differentiates our capability is that we do have all of the building blocks of CPUs, GPUs, as well as all of the networking technologies that you would need to put the solutions together.”
Conclusion:
What will get us to close our position? Price, first and foremost – which is why we sizably trimmed the position a few months ago. From what I’m hearing on the technicals side, price is doing what was expected, but you will know if this changes via our real-time trade alerts. Secondly, if the MI350s are too weak to take the overflow in capex that Blackwell supply cannot fulfill. If consumer and tariffs create more impact than currently priced in. We always have a plan.
With that said, I will be watching with keen interest the launch of the MI350s. If this release benchmarks “close enough” to the B200s then AMD investors will be rewarded. If the benchmarks are too far apart, the proverbial David (Lisa Su and team) will have missed a critical window for striking with her slingshot. Despite what benchmarks say, we only have to watch how Big Tech capex funnels to truly know how AI accelerators are being received; this is a proxy and substantial (unprecedented) flow of capital that overwhelmingly matters more than what analysts, industry experts, or even management teams say or predict. Big Tech companies are at the forefront and they will make the ultimate decision.
Right now, that capex is pointed directly at Nvidia, which was the basis of our Q1 webinar. We will see come late Spring/early Summer if AMD can pry some of the roughly $300 billion loose. The market is giving up, but that sometimes becomes precisely the moment when a stock bottoms.
Before I conclude, keep an eye on the Client segment as it’s abundantly clear who is in the lead with AI PCs. Nvidia will do quite well here too, but AMD’s numbers are communicating something very important about their lead in edge AI.
Royston Roche, Equity Analyst for the I/O Fund, contributed to this analysis.
Please note: The I/O Fund conducts research and draws conclusions for the company’s portfolio. We then share that information with our readers and offer real-time trade notifications. This is not a guarantee of a stock’s performance and it is not financial advice. Please consult your personal financial advisor before buying any stock in the companies mentioned in this analysis. Beth Kindig and the I/O Fund own shares in AMD at the time of writing and may own stocks pictured in the charts.
Recommended Reading: