Perplexity Picks Nvidia Over AMD: What It Means for NVDA and AMD Stock

What Perplexity's Chip Choice Changed

Perplexity, the fast-growing AI search startup, has reportedly built its infrastructure around Nvidia chips rather than Advanced Micro Devices hardware, according to a report examining how AI companies are picking sides in the chip race. On its own, one customer's infrastructure decision does not move either company's revenue in a meaningful way. What makes it worth noting is the pattern behind it: years into the generative AI boom, fast-scaling AI companies still default to Nvidia's chips and software stack almost automatically, rather than treating chip suppliers as interchangeable commodities.

Why Nvidia Stock Is in Focus

Nvidia's data-center GPUs, paired with its CUDA software layer, remain the path of least resistance for any company that needs to train or run large AI models quickly. For Nvidia, the earnings effect of any single AI startup's hardware choice is small, since its data-center business already counts nearly every major AI lab, cloud provider and startup as a customer. The value of stories like this one is reputational rather than financial: each publicized win reinforces the idea that switching away from Nvidia carries real technical and time cost, which supports the premium the market already assigns to its AI franchise.

Why AMD Stock Faces Pressure

For AMD, being passed over by a prominent AI company is a familiar story rather than a new setback. AMD's Instinct MI-series accelerators have closed real ground on raw performance and have landed a growing list of large customers, but reports like this one show the harder part of the fight: convincing newer, fast-moving AI companies to build their core infrastructure around AMD's chips and software instead of the more established Nvidia stack, where engineering teams already have deep experience. The direct hit to AMD's own numbers from one startup's decision is limited, since AMD's revenue spans CPUs, gaming chips, embedded processors and a still-growing AI accelerator line, but each story like this makes it a little harder for AMD to win over the next customer.

What to Watch Next

The more meaningful signal for both stocks will come from AMD's own quarterly disclosures on AI accelerator revenue and any new hyperscaler design wins, set against Nvidia's data-center revenue growth in its coming reports. Watch specifically for whether AMD's newer MI350 and MI400 series chips start appearing in announced deployments at AI-native companies rather than only at large cloud providers, since that would be the clearer sign that the competitive gap is actually narrowing. A widening list of such wins for either company over the next few quarters would also test how much pricing power Nvidia can keep as more AI compute capacity comes online industry-wide.