Amazon and Nvidia have announced a major expansion of their partnership, with Amazon significantly increasing its order for Nvidia GPU chips. This strategic alliance highlights the intense demand for advanced computing power in the rapidly evolving field of artificial intelligence.
Amazon Bolsters AI Infrastructure with Massive Nvidia Chip Order
In a significant development revealed on Wednesday, August 26, 2026, during Nvidia's quarterly earnings call, Amazon and Nvidia unveiled a substantial amplification of their existing collaboration. Amazon is set to acquire an additional 2 million Nvidia GPU chips for deployment in its Amazon Web Services (AWS) data centers over the next two years, specifically in 2027 and 2028. These high-performance GPUs, encompassing Nvidia's cutting-edge Blackwell Ultra, Rubin, and Rubin Ultra models, are crucial for managing the intensive computational demands of AI model training and operation. This latest order comes merely five months after Amazon initially committed to integrating over 1 million Nvidia GPUs into its AWS infrastructure starting in the current year, a plan that saw "demand exceed expectations."
While specific financial details of the expanded agreement remain undisclosed by both parties, industry analysts estimate the deal to be worth tens of billions of dollars, given the high unit cost of these specialized GPUs. The partnership's scope extends beyond merely purchasing more hardware. Nvidia confirmed that its comprehensive technology stack, including networking hardware designed to interconnect thousands of GPUs, open models, CPUs, data processing software, and its advanced robotics platform, will also be integrated throughout AWS. This broad integration underscores a deepening strategic alignment between the two tech giants. Both companies cite "surging demand" from a diverse client base—ranging from startups and large enterprises to AI research laboratories and even governmental entities—as the primary catalyst for this intensified collaboration.
This expansion is particularly noteworthy as Amazon continues to invest heavily in developing its own artificial intelligence chips, including its Trainium chips, which directly compete with Nvidia's H100 and Blackwell chips for deep learning tasks, and its Arm-based Graviton CPUs, challenging traditional server chip manufacturers like Intel and AMD. Despite Amazon's growing custom chip business, which recently surpassed a $25 billion annualized revenue run rate driven by significant commitments from AI labs such as Anthropic and OpenAI, Nvidia remains a dominant force in the AI chip market. Nvidia's CFO, Colette Kress, further elaborated that an unspecified number of Vera CPUs would accompany the 2 million GPUs destined for AWS, with some integrated into Rubin units and others operating independently. Nvidia CEO Jensen Huang has expressed ambitious projections for the Vera CPUs, envisioning a "brand new $200 billion TAM" for the company. Kress also noted that Vera CPUs are expected to be adopted by "every major hyperscaler, neocloud, AI lab, and system OEM," with initial shipments already underway to lead partners like Oracle and SpaceXAI.
The collaboration also extends into Amazon's physical operations and enterprise solutions. Amazon plans to adopt Nvidia's complete physical AI stack to enhance its fleet of warehouse robots. This stack includes Omniverse for simulation and digital twin creation, Cosmos for world modeling, Isaac for robotics development, and Jetson for edge AI and robotics computing hardware. Nvidia recently unveiled a new version of Jetson, designed to make entry-level edge AI more accessible. On the enterprise front, AWS will host Nvidia's Nemotron family of open models on its managed foundation model platform, Amazon Bedrock, and its managed cloud service, SageMaker. Financially, Nvidia reported robust second-quarter sales of $96.2 billion, exceeding analyst forecasts, with data center revenue accounting for $89 billion—a 117% increase year-over-year. The company projects third-quarter revenue to reach $108 billion, buoyed by sales of its next-generation Rubin GPUs, which entered production shipments this quarter. Nvidia's substantial commitment of $279 billion to secure supply and manufacturing capacity for upcoming data center projects, a significant jump from $119 billion last quarter, underscores its strategy to meet the sustained demand for AI hardware over the coming years.
This deepening partnership between Amazon and Nvidia signifies a pivotal moment in the AI industry. It underscores the immense and rapidly growing demand for specialized AI computing power, even as major tech players like Amazon simultaneously pursue their own chip development. The integration of Nvidia's comprehensive AI ecosystem, from advanced GPUs to robotics platforms and enterprise AI models, into Amazon's vast infrastructure points towards a future where AI-driven services become even more ubiquitous and sophisticated. It also highlights the strategic imperative for companies to secure foundational AI technologies, suggesting that the "AI gold rush" is far from over, with substantial investments continuing to pour into both hardware and software innovation to capitalize on the transformative potential of artificial intelligence.