For investors in the technology sector, generative artificial intelligence has become a “gift” that continues to deliver returns, and chip manufacturer Advanced Micro Devices (AMD) is among the biggest beneficiaries. However, after its stock price surged over 300% in the past twelve months, the key question for the market is whether this stock can sustain its upward momentum.
The training and operation of large language models typically rely on massive data centers equipped with thousands of graphics processing units (GPUs) and dedicated AI accelerator chips—hardware that AMD produces. Currently, there are no signs of slowing market demand.
Industry leaders widely believe that AI will transform the world. For example, Amazon CEO Andy Jassy has called this technology a “once-in-a-generation” opportunity. According to media reports, major cloud providers are increasing their investments, with total spending potentially reaching $700 billion this year. This situation has created strong demand and relatively tight supply, significantly boosting AMD’s operating performance.
In the first quarter of this year, AMD’s revenue grew 38% year-over-year to $10.25 billion, primarily driven by the strong performance of its data center business. Revenue from this segment rose 57% to $5.8 billion. AMD’s EPYC processors (central processing units that power cloud computing) have succeeded, while the continued ramp-up of its Instinct GPUs (used for AI model training and inference) has fueled investor expectations for subsequent growth momentum.
The company’s gross margin improved by 300 basis points to 53%, and operating income grew 83% to $1.48 billion.
AMD’s business diversification is a key advantage over its competitors. Unlike Nvidia, which derives over 90% of its sales from the data center business, AMD’s data center segment accounts for only 57% of its total revenue. Furthermore, within this business segment, AMD sells a healthy mix of GPUs, CPUs, and other networking equipment, reducing dependence on any single product category or customer group.
However, AMD’s business model also faces challenges. The company’s first-quarter gross margin was 53%, significantly lower than Nvidia’s 74.9%. This lower-margin product mix may put it at a disadvantage when competing with rivals.
AMD also faces macroeconomic uncertainties related to the entire AI industry. It has been four years since OpenAI launched ChatGPT, and large language models remain interesting novelties, far from the life-changing productivity tools many tech executives have promised. Many companies pioneering this technology have yet to achieve profitability. Deutsche Bank analysts estimate that by 2029, OpenAI alone will have cumulative operating losses and capital expenditures totaling $140 billion.
Although AMD, as an infrastructure provider, can somewhat mitigate the risks and uncertainties faced by consumer-facing software companies, a slowdown in the development of large language models could ultimately reduce demand for its data center hardware, severely weighing on the company’s stock price.
Currently, AMD’s stock trades at a forward price-to-earnings ratio of 74, far above the Nasdaq 100 Index’s average expected valuation of 27, making the stock appear overly expensive. While the company’s diversification is attractive, its gross margin is significantly lower than that of other AI infrastructure alternatives (such as Nvidia). AI stocks have delivered substantial returns for early investors, but this opportunity no longer seems fresh. Investors may be better off staying on the sidelines, waiting for more information to judge whether current data center spending levels are sustainable over the long term. And the growing likelihood suggests that such high spending levels may be difficult to maintain.
