Amazon and Nvidia announced an expanded partnership Wednesday that will add another 2 million Nvidia GPU chips to Amazon’s data centers, a deal that arrives just five months after Amazon agreed to deploy more than 1 million Nvidia GPUs across AWS infrastructure. Nvidia said in a statement that “demand has exceeded those expectations” since that earlier agreement.
What’s in the Deal
The new GPUs, including Nvidia’s Blackwell Ultra, Rubin, and Rubin Ultra chips, will be delivered to AWS data centers in 2027 and 2028. Neither company disclosed financial terms, but given the cost of GPU units at this scale, the deal is worth tens of billions of dollars.
The partnership extends well beyond chip purchases. Nvidia said its networking hardware, which connects thousands of GPUs into a single system, along with its open models, CPUs, data processing software, and robotics platform, will also be integrated across AWS. Nvidia CFO Colette Kress said the company will also send an unspecified number of Vera CPUs to Amazon, “some integrated with Rubin, others standalone.”
Why Now
The companies cited “surging demand” from startups, enterprises, AI labs, and governments as the driver behind the expanded partnership. The announcement came during Nvidia’s quarterly earnings call, where the company reported $96.2 billion in second-quarter sales, beating analyst estimates, with data center revenue alone hitting $89 billion, up 117% year-over-year. Nvidia guided to $108 billion in third-quarter revenue, with some of that growth expected from its next-generation Rubin GPUs, which began production shipments this quarter.
Nvidia CEO Jensen Huang described the underlying economics driving the buildout: “AI is generating profitable tokens… If we had more compute, we could generate more profitable tokens, which results in more profit for all of the services. This is the exact phase where we’re at, which is the reason why everybody’s leaning in.”
A Complicated Relationship
The deal is notable partly because it deepens even as Amazon simultaneously invests in chips designed to compete with Nvidia. Amazon’s AI chief Peter DeSantis has said AWS is in talks to sell its Trainium chips, direct alternatives to Nvidia’s H100 or Blackwell hardware, to other companies for use in their own data centers. Amazon’s Arm-built Graviton CPU is also positioned as a challenger to traditional server chips from Intel and AMD. Amazon has said its custom chip business crossed a $25 billion annualized revenue run rate, backed by $225 billion in total commitments from AI labs including Anthropic and OpenAI.
Despite that competing effort, Amazon is still expanding its Nvidia footprint substantially, an indication that even companies building their own chip alternatives don’t see them as a near-term replacement for Nvidia’s most advanced hardware.
Beyond Data Centers
The expanded partnership also extends into physical robotics. Kress said Amazon plans to adopt Nvidia’s full physical AI stack, Omniverse (simulation and digital twins), Cosmos (world models), Isaac (robotics development), and Jetson (edge AI computing hardware), to power its warehouse robot fleet. Nvidia this week also introduced a new version of Jetson aimed at “entry-level edge AI.” On the enterprise side, AWS will host Nvidia’s Nemotron family of open models on Amazon Bedrock and SageMaker.
Kress added that Nvidia’s Vera platform is expected to be deployed by “every major hyperscaler, neocloud, AI lab, and system OEM,” with shipments already underway to lead partners including Oracle and SpaceX AI.
Why It Matters
Nvidia has now committed $279 billion to secure supply and manufacturing capacity for current and future data center projects, up sharply from $119 billion the previous quarter, as it races to lock in memory and manufacturing capacity ahead of continued AI demand. For Amazon, the expanded order signals that internal chip development, however fast it scales, isn’t yet reducing its reliance on Nvidia’s top-end hardware. For the broader industry, the scale of the commitment, and how quickly it grew beyond initial expectations, is one of the clearest signals yet that AI infrastructure spending is still accelerating rather than plateauing.
For continuing coverage of AI chip supply deals and cloud infrastructure investment, keep following Tech News Reports for ongoing updates.
