Amazon chief executive Andy Jassy used the company’s first-quarter 2026 earnings call to deliver a message that Nvidia (NASDAQ: NVDA) shareholders will have welcomed: despite Amazon’s rapid expansion into its own chip development, the relationship between the two companies remains deep, durable, and commercially significant for the foreseeable future.
Amazon has been building out its proprietary silicon business at a pace that has attracted significant attention from the investment community. The company’s chip revenue rose 40% sequentially in the first quarter of 2026, with the run rate growing at a triple-digit percentage rate.
That trajectory has positioned Amazon as one of the three largest data centre chip companies in the world on a standalone basis, a remarkable development for an organisation whose primary identity remains cloud infrastructure rather than semiconductor design.
The in-house chips at the centre of that story are Trainium, designed for machine learning training workloads, Graviton, which handles general compute, and Nitro, which underpins the virtualisation layer of Amazon Web Services. These chips serve a specific purpose within Amazon’s broader strategy: offering budget-friendly options for the vast population of smaller developers and businesses on the AWS platform that do not need the most powerful hardware available. For that segment of the market, Amazon’s custom silicon is an effective and cost-competitive alternative.
But for the companies building the most demanding and sophisticated AI applications, the picture is different. When Jassy addressed the Nvidia relationship directly on the earnings call, his language left little room for ambiguity. “While the largest number of AI chips we are bringing in are Trainium, we continue to have a deep partnership with Nvidia,” he said. “We have immense respect for them, continue to order substantial quantities, will be partners for as long as I can foresee, and we will always have customers who want to run Nvidia on AWS.”
The commercial context for that statement is significant. Amazon’s AI business is growing at triple-digit percentage rates and signed major new agreements in the first quarter alone with OpenAI, Meta Platforms, Nvidia itself, Southwest Airlines, the US Army, and Bloomberg. When enterprise clients of that calibre build generative AI applications at scale, they demand the most powerful chips available, and Nvidia’s GPU architecture remains the dominant choice for that category of workload. Amazon’s custom chips are competitive for many use cases but are not a wholesale replacement for Nvidia’s top-tier hardware, a reality that Jassy effectively acknowledged in his comments.
The timing of his remarks matters for Nvidia investors heading into what is expected to be a landmark earnings report. Wall Street is projecting Nvidia to report approximately $79 billion in revenue for the first quarter of its fiscal 2027, which would represent a 79% increase year over year and an acceleration from the 73% growth recorded in the prior quarter. Nvidia has a consistent history of exceeding analyst expectations, and the confirmation from one of its largest customers that substantial orders will continue arriving provides additional confidence ahead of the May 20 results.
Nvidia’s stock (NASDAQ: NVDA) has recovered from a dip earlier in 2026 and was up approximately 6% year to date as of the date of Amazon’s earnings call, broadly in line with the wider market. However, the stock remains approximately 8% below the peak it reached in 2025, leaving room for upside if the Q1 fiscal 2027 results deliver the kind of blowout numbers the market has come to expect from Jensen Huang’s company.
For investors in both names, Jassy’s remarks capture the most important dynamic in AI infrastructure right now: even the companies building their own chips at scale still need Nvidia, and that need is growing rather than diminishing as the workloads being run on cloud platforms become more complex and more resource-intensive. Amazon’s chip business is a genuine competitive development, but Nvidia’s position at the top of the AI hardware stack appears to be one that no customer, however large, is currently in a position to replicate.
Ewan is a reporter at the Foreign Policy Journal, primarily covering US current affairs and news related to NYSE-listed stocks. He has over two decades of experience in journalism and digital publishing.
