David Becker
Investment Thesis
Since I last covered Advanced Micro Devices, Inc. (NASDAQ:AMD) in April, the stock has barely outstripped the market, increasing by 13.49%, while the S&P 500 (SP500) increased by 10.33%. Looking at their biggest competitor, NVIDIA Corporation (NVDA), this company increased by roughly 50% within the same timeframe.
Despite (what I believe to be is) their 2nd place rating in the AI-chip market, AMD has obviously not been able to keep up with Nvidia’s share price. Although they outperformed the market, I believe their stock should have increased at a rate more like their competitor. I suspect this is due to the belief held by some investors that Nvidia has a runaway lead in the AI GPU race. However, there is so much more that deserves to be factored in than just GPUs and CUDA.
AMD understands that the AI chip race will be centered around the combination of CPUs on computers, and the use of multiple GPUs and servers, not just enterprise GPUs. Nvidia technology can really only operate well with other Nvidia technology (hence CUDA), and AMD does not think this centralization around one company is the future of AI. They believe that it is an ecosystem, with a collection of technology from multiple companies that creates the bedrock of future AI model developments. However, the market has not yet come to this same conclusion, and I believe this is a possible reason for their undervaluation.
This undervaluation can be seen in their valuation metrics. The forward PEG ratio for AMD stands at 1.16, significantly lower than the sector median of 2.06, indicating that the market may not fully appreciate AMD’s growth potential.
To build an edge in the AI market and develop an AI ecosystem, AMD recently acquired Silo AI for $665 million. This acquisition integrates advanced AI technologies, such as large language model training and AI model development, into AMD’s product lineup. This move strengthens AMD’s AI production and positions them to better compete with Nvidia’s dominance in the AI market.
While AMD’s stock has underperformed relative to Nvidia and barely beat the market, I believe the company’s focus on building a comprehensive AI ecosystem is the key to success. Once the market realizes this, I believe we will see dramatic growth not only in operating results but also in shares. With this, I still believe AMD is a strong buy.
Why I Am Doing Follow-Up Coverage
As I mentioned above, since my coverage of AMD in April, the company has barely outstripped the market. Considering AMD is the second-best AI chip producer, this just doesn’t seem right.
It is clear the market is doubting their future earning potential as their PEG ratio is well below the sector median (they doubt once their true earnings results come out they will perform below forward expectations). I believe this is misguided and overly bearish. With this, I think follow-up coverage is necessary. The purpose of this article is to explain how AMD is constructing an AI ecosystem that the market doesn’t yet see.
AI Ecosystem: What Does This Mean?
AMD’s approach to the AI revolution goes beyond simply developing powerful GPUs for enterprise servers. The company’s strategy entails creating a diverse range of AI-enabled hardware, including GPUs, CPUs, and other processors, designed to cater to both personal work computers and servers. This holistic approach is indicative of AMD’s understanding that the future of AI lies in an integrated ecosystem rather than concentrated components on cloud mainframes. With this, there is a growing trend that companies are seeking more control over their computation process (more on-premise compute), and this means that they are looking for technology that they can control locally from employee’s computers and servers. AMD can help fit this increasing demand.
With this, AMD has been working on a collection number of projects that knit into this ecosystem. For example, during the AMD Investor Conference, Jean Hu, AMD’s CFO, mentioned one of their new products stating:
…the first is AIPC, because it is a consumer show we actually announced about our right in AIPC 300 series. It’s for premium, like actual thing notebook. It actually has the GP and the latest of the GPU and the NPU, which are all on one single chip. The top Microsoft one is 40 pops — we actually have 20% more tops, we are the only one who reached the 50 pops to really power the Copilot plus the AIPC. So it’s one of the really amazing product. It’s going to be available in July – Investor Conference.
This is a great combination of compute. AMD is rising above the current industry standards with Microsoft Corporation (MSFT) AI PCs. This is powerful.
She then went on to discuss various new products in the works:
We also announced our Ryzen 9000 desktop processors. It also has leadership of performance in AI inferencing. So that’s on the PC side. Then on the data center side, we actually previewed our next-generation Gen 5 EPYC CPU servers, codenamed Turin, to extend our leadership in both performance per watt and performance per dollar significantly – Investor Conference.
I believe it is clear that AMD is developing new technology at a rapid pace, which is critical in the AI market. They are producing various technology innovations, not just one like some of their competitors, hitting both major markets, PC, and server. The company acknowledged this during the conference, with Jean Hu stating:
AMD is the only company who have end-to-end solutions covering the CPU, GPU, and MPU from a data center and then to PC and eventually our embedded business, which is Xilinx FPGA business, we do think edge AI will happen in the future. So we have a broad portfolio to cover everything – Investor Conference.
I believe if the trend continues of companies wanting more control and access to their AI systems (more on-premise servers), they will look towards AMD for help. Back in February, I mentioned AMD’s 2.5D/3D technology. This technology will be key here, as it allows AMD to incorporate any chiplets into one package, increasing performance and space optimization while decreasing power consumption. This will help them build the best Ecosystem of AI chips.
AMD’s acquisition of Silo AI for $665 million adds to this. The startup helps them with their rapid development of new technology. For instance:
[this acquisition will allow AMD access to Silo’s renowned] team of AI scientists and engineers who specialize in creating customized AI models and solutions whose expertise spans various sectors, including cloud computing and embedded systems -Press Release.
In addition to this, AMD “will now be able to offer not only hardware, but also software for enterprise customers.” As I mentioned before, software is a considerable part of this AI ecosystem. This acquisition helps fill that.
AMD acknowledged the benefits of this acquisition during a statement where senior vice president, Vamsi Boppana, stated:
Across every industry, enterprises are looking for fast and effective ways to develop and deploy AI solutions for their unique business needs…silo AI’s team of trusted AI experts and proven experience developing leadership AI models and solutions, including state-of-the-art LLMs built on AMD platforms, will further accelerate our AI strategy and advance the build-out and rapid implementation of AI solutions for our global customers. -Silo AI Acquisition Press Release.
Having access to Silo’s expertise and solution will be huge for AMD as it will open new markets for them to address, helping them attempt to gain a competitive edge over competitors like Nvidia.
Earnings Expectations Are Coming Down
While this article is not meant to be an earnings preview, the market is downgrading their earnings expectations, despite what I believe to be a clear notion that the AI revolution is accelerating in many aspects.
Over the past three months, there have been 30 downward revisions for AMD’s forward EPS estimates, compared to only 9 upward revisions. The consensus EPS estimates for AMD show a mixed picture, with a slight decline (on a 1-month basis) in the near term (December 2024) to $3.50/share, but significant growth expected in the coming years, with the EPS reaching $7.30 by 2026.
Revenue estimates also reflect this cautious outlook. There have been 28 downward revisions compared to 13 upward. The consensus revenue estimate for December 2024 is $25.54 billion, marking a 12.60% year-over-year growth. In this case, the estimate has decreased by 0.08% in the past month and 1.04% over the past six months. Despite these downward revisions, the longer-term revenue growth projections remain robust, with significant increases expected through 2026 to $38.70 billion.
As I mentioned above, this cautious market sentiment comes at a time when the AI revolution is accelerating in many aspects. During the AMDs session at the BofA Securities Conference, CFO Jean Hu stated:
the reason we are accelerating our roadmap is because we see the demand continue to exceed our expectations. We see customers need two suppliers for this very, very large market – BofA Conference.
This literally runs contrary to the concept of downgrading forward EPS estimates. Demand is beating even the biggest optimists’ expectations (the company insiders).
Despite what, I believe, is obvious heightened demand for AMD’s technology, the market isn’t pricing this in. I think these earnings revisions are very misleading of the company’s current performance. It represents a unique opportunity, however.
Valuation
I believe one of the key valuation metrics to consider for the chipmaker is their forward non-GAAP PEG ratio. Their metric currently stands at 1.16 which is 43.50% below the sector median of 2.06. I mentioned previously how this doesn’t match up with the growth of the AI market. Based on the quotes from the BofA conference, the Investors conference, and the acquisition of Silo, I think this is a severe undervaluation. I strongly believe earnings expectations are moving in the wrong direction, which only means their forward PEG ratio may be even lower.
To this point, their below-sector median PEG ratio just doesn’t make sense to me given their EPS growth. AMD’s forward EPS GAAP growth is 59.04%, 716.08% higher than the sector median of 7.23%. The projections for EPS growth support this narrative as well, considering that by 2028, the EPS is supposed to hit $10.78/share, growing from the predicted December 2024 value of $3.50. The difference between these two EPS values, in my opinion, is a massive disparity given the advancements in technology coming from AMD and high demand.
If AMD’s PEG were to converge on the sector median, this would represent roughly a 77.6% upside. Once the market realizes AI is an ecosystem, I think we will see massive growth in operating performance (and hopefully in share price)
Risks
I believe AMD’s growth prospects are promising, but it’s not a clean-cut story.
The biggest concern of mine is the potential slowing of AI demand in some pockets, which could affect AMD’s growth. Some banks like Goldman Sachs have called attention to the AI trade, concerned that the price action has largely gotten ahead of itself.
With this, the competitive landscape, of course, also poses a risk. Despite their ecosystem advancements, AMD is still underperforming industry giant Nvidia for market share. Nvidia holds a dominant position in the AI GPU market with their CUDA-based technology, which allows software to link a series of GPUs together so they can perform computations in sync and allows them to perform computations faster together. AMD’s acquisition of Silo AI is a strategic move to challenge Nvidia’s dominance, but integrating and leveraging this acquisition effectively will be crucial.
Despite this, I’m not concerned. I think the market’s outlook on AMD actually remains too bearish and a lot of the Goldman AI concerns are baked in. Not only because of the undervaluation when it comes to their EPS estimates, but also due to AMD barely outperforming the market and their below-sector median PEG ratio. I think there is a real potential that AMD will surprise to the upside.
Bottom Line
While the company has underperformed their biggest competitor Nvidia and barely beat the market since my last report, I continue to be firm that I just don’t think the market has arrived at the same conclusion: AMD is making more than just chips, they are making an AI ecosystem.
While there is a high P/E risk (of course), and risk of slowing growth, I am not concerned. Based on the qualitative projections AMD has announced, I am confident they will not only meet, but beat the market’s expectations. I continue to believe AMD is a strong buy.
Credit: Source link
