With shares up a whopping 3,300% over the last five years, Nvidia (NVDA +3.04%) is winning the artificial intelligence (AI) race by a landslide. For comparison, rival Advanced Micro Devices (AMD 0.80%) is only up by just under 500% in the same time frame.
As Nvidia soars to higher and higher valuations, investors may want to look for more under-the-radar ways to bet on the long-term AI hardware opportunity. Could AMD fit the bill? Let's dig deeper into the potential advantages this smaller chipmaker could have over the market leader.
NASDAQ: AMD
Key Data Points
Aiming to capture market share
Since it acquired ATI technologies in 2006, AMD has been locked in a battle with Nvidia in the market for graphics processing units (GPUs), a type of computer chip that excels at performing complex tasks like rendering video game visuals, cryptocurrency mining, and training AI algorithms. Historically, Nvidia has dominated the high end of the opportunity, forcing AMD to compete based on pricing and value.
This trend holds true in AI chips, where Nvidia is believed to control 70% to 95% of the market for GPUs used to train large language models (LLMs) like OpenAI's ChatGPT or Alphabet's Google Gemini.
AMD is not ignoring this long-term opportunity, which management believes could be worth a whopping $400 billion by 2027. This month, the company announced its latest AI chip, the M1325X, designed to perform 35 times better in inference (running AI models) compared to its predecessor, the Mi300. Nvidia is updating its AI chips once per year, and AMD will probably have to keep up the pace to remain relevant.
What advantages does AMD have over Nvidia?
AMD has some advantages over Nvidia. According to industry magazine Tom's Hardware, the company's current flagship AI GPU (the Mi300x) costs just $10,000 to $15,000 per unit compared to Nvidia's H100, which can go for above $40,000. And although the industry leader has wide a economic moat because of its software solution CUDA, AMD could attract smaller customers who simply can't afford Nvidia's prices or want to avoid becoming too dependent on one hardware supplier.
Image source: Getty Images.
From an investor perspective, AMD's business diversification is also attractive relative to its rival. As of the first quarter, Nvidia generated 87% of its sales from data center chips, which means it is extremely overexposed to any potential demand slowdown in this market.
On the other hand, data center chips only represented 42% of AMD's first-quarter revenue, so it doesn't have all its eggs in one basket. But to be fair, this dynamic is mostly because Nvidia's data center business has grown significantly faster than AMD's over the last two years.
Which stock is the better buy?
Despite its advantages, it's hard to call AMD a better buy than Nvidia for one big reason: valuation. Despite a higher market cap, Nvidia's forward price-to-earnings (P/E) multiple of just 44 means it is cheaper than AMD (forward P/E of 48) when considering bottom-line growth rates.
With that being said, Nvidia's overreliance on data center GPU sales could become a significant risk factor if the industry slows down. So while AMD is not a better pick for all investors, it could be ideal for those who want exposure to long-term AI opportunity without going all in.
