Can Amazon Take Market Share From Nvidia in the AI Chip Race?

Amazon (NASDAQ: AMZN) recently launched a new software tool that will allow its custom Trainium AI chips to run code written in the most popular AI coding language without modification. It's a move aimed at boosting the potential for Amazon's custom application-specific integrated circuits (ASICs) to take market share away from Nvidia's (NASDAQ: NVDA) widely used graphics processing units as demand for AI chips continues to swell.

Amazon isn't the only tech giant working on its own AI chips for commercial use. Alphabet recently began making its custom AI chips available for sale directly to enterprises, rather than limiting them to being leased through Google Cloud. However, investors shouldn't expect these developments to have a meaningful effect on Nvidia's financial results.

Image source: Getty Images.

Amazon has a relatively small chip business

Amazon's AI chip business wrapped up the second quarter with a $25 billion annual revenue run rate. That's not small, but it pales in comparison to the $96.2 billion in revenue Nvidia booked in its fiscal 2027 second quarter alone, or the more than $300 billion it booked over the past four quarters.

While Amazon's push to commercialize its custom chips could boost sales, the largest tech companies are still rushing to buy Nvidia chips. That's how Nvidia managed to produce 106% year-over-year revenue growth in its most recent quarter, and CEO Jensen Huang says that AI demand is still accelerating.

If anything, Amazon's AI chips can become niche hardware in the industry. Since the AI build-out has been so expansive, it may provide enough room for high, sustained revenue growth rates. If Amazon's chip business can generate net profit margins similar to Nvidia's, that should be enough to boost Amazon's stock price.

Amazon is still buying Nvidia chips

Another point supporting the view that Amazon's AI chips won't significantly impact Nvidia's business is that Amazon is still buying Nvidia chips. In fact, it has built up a stockpile of billions of dollars' worth of them that it doesn't yet have the data center infrastructure and connected electricity supply to power up.

Nvidia's processors have become the most reliable option for providing high-end compute in an industry that values speed and will pay a premium for flawless execution. Even as Amazon touts the value of its own chips for the narrower workloads they are designed to handle, for flexible compute power, the tech leader still turns to Nvidia.

Notably, those stockpiled supplies of chips and data center servers highlight that electrical supplies will remain a major bottleneck for AI infrastructure expansion. Still, hyperscalers aren't waiting for that problem to be resolved completely, and continue to accelerate their capital expenditures.

Nvidia's guidance implies that its business is still growing at a substantial rate despite the intensifying competition. Its fiscal 2027 third-quarter guidance calls for $108 billion in revenue at the midpoint, which would amount to 12.3% sequential growth. That guidance supports Huang's claim that AI demand is accelerating.

While more rivals are likely to enter the expanding AI chip market, Nvidia remains the clear leader in data center chips.

Marc Guberti has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Alphabet, Amazon, and Nvidia. The Motley Fool has a disclosure policy.