AWS and NVIDIA Expand Partnership with 2 Million More GPUs

Amazon Web Services and NVIDIA announced an expansion of their strategic collaboration on August 27, 2026, with a plan to deploy an additional 2 million GPUs for agentic and physical AI workloads. The announcement, which signals a massive scaling of AI infrastructure capacity, preceded a modest early-market response in NVIDIA's stock price. As of 05:53 UTC today, NVIDIA shares traded at $209.66, a gain of 0.57% on the day. The stock had reached an intraday high of $213.60 before settling into a narrow range between $209.23 and the session peak. This partnership extension underscores the intensifying race among cloud providers to secure next-generation AI computing power.

Context — why this matters now

The scale of this GPU commitment, numbering in the millions of units, is unprecedented in the history of cloud computing. In May 2024, NVIDIA and AWS announced a previous expansion involving hundreds of thousands of GPUs, which was considered a landmark deal at the time. The new agreement represents a nearly order-of-magnitude increase in planned capacity, reflecting the explosive demand for AI model training and inference. The partnership focuses specifically on agentic AI, which involves autonomous systems that can perform complex tasks, and physical AI, which bridges digital intelligence with real-world robotics and sensors.

The current macro backdrop for technology investments remains strong, with the Nasdaq Composite index near all-time highs driven by AI-related earnings. Long-term Treasury yields have stabilized, providing a favorable environment for capital-intensive infrastructure projects. The timing of this announcement is critical as enterprises move from piloting generative AI applications to deploying them at scale in production environments. This requires a significant and reliable supply of high-performance computing resources that this partnership aims to provide.

A key catalyst for this expansion is the competitive dynamics in the cloud sector. Microsoft Azure's deep partnership with OpenAI and Google Cloud's investments in its Tensor Processing Units have created intense pressure on AWS to demonstrate its AI capabilities. Securing a GPU allocation of this magnitude from NVIDIA, the dominant supplier of AI accelerators, is a strategic counter-move. It ensures AWS can compete for large-scale AI workloads from major corporations and hyperscalers.

The AI hardware market is also transitioning towards more specialized workloads. Agentic AI requires low-latency, high-throughput systems for real-time decision-making. Physical AI demands integration with control systems and edge computing networks. This partnership likely involves NVIDIA's newest Blackwell architecture GPUs, which are optimized for these specific use cases. The deployment will occur over multiple years, aligning with NVIDIA's production roadmap and AWS's data center expansion plans.

Data — what the numbers show

The market's initial reaction to the partnership news was measured. NVIDIA's stock gained $1.19, or 0.57%, to trade at $209.66. The stock's trading range for the session was relatively tight at just over $4, between $209.23 and $213.60. This suggests that while the news was positively received, it may have been partially anticipated by investors given the companies' long-standing relationship. The magnitude of the gain is modest compared to some of NVIDIA's larger single-day moves driven by earnings surprises.

A comparison of recent major NVIDIA announcements and subsequent stock performance reveals a pattern. The May 2024 AWS partnership announcement coincided with a 3.4% single-day gain. A February 2025 earnings beat that exceeded revenue estimates by 12% triggered a 7.8% surge. The 0.57% move following this latest news places it on the lower end of the reactivity scale for significant partnership announcements. This could indicate that the market had priced in expectations for continued cloud expansion or is awaiting more specific financial details.

The scale of the GPU commitment itself is the most striking data point. Two million GPUs represent a substantial portion of NVIDIA's projected production capacity. For context, analysts estimate NVIDIA shipped approximately 3.7 million data center GPUs for the entire fiscal year 2025. While this new AWS commitment will be fulfilled over multiple years, it locks in demand for a significant volume of future output. Assuming an average selling price of $30,000 per high-end data center GPU, the deal could represent a theoretical value of $60 billion over its lifespan, though actual pricing for large-scale contracts is typically lower.

The announcement's impact on related equities was mixed. Advanced Micro Devices, NVIDIA's primary competitor in the AI accelerator market, saw its shares decline slightly in pre-market trading. Cloud software companies that are heavy users of AI infrastructure, such as Salesforce and Adobe, traded flat to slightly positive. The lack of a broad-sector rally suggests investors are carefully evaluating the balance between increased infrastructure supply and the demand required to utilize it profitably. The deal solidifies NVIDIA's market leadership but also highlights the immense capital expenditure required to compete in the AI platform race.

Metric	Value	Context
NVIDIA Stock Price	$209.66	As of 05:53 UTC, August 27
Daily Gain	+0.57%	+$1.19
Intraday High	$213.60	Session peak
GPUs Committed	2,000,000	For agentic and physical AI

Analysis — what it means for markets / sectors / tickers

The primary beneficiary of this expanded partnership is clearly NVIDIA. The company secures a massive, multi-year commitment for its flagship products, providing revenue visibility and justifying its continued research and development investments. This deal likely involves NVIDIA's most advanced GPUs, which carry higher margins than previous generations. It also strengthens NVIDIA's ecosystem lock-in, as developers building on AWS's AI platform will be optimizing for NVIDIA's CUDA software stack, creating a durable competitive advantage.

For AWS, the investment is a necessary defensive and offensive move. It ensures the cloud leader remains competitive in the AI infrastructure arms race against Microsoft and Google. However, the capital expenditure required for this deployment will pressure AWS's operating margins in the short to medium term. The profitability of this investment hinges on achieving high utilization rates for these GPUs, which depends on sustaining the current wave of enterprise AI adoption. AWS will likely bundle these powerful resources with its own AI software services to create higher-value packages for customers.

Second-order effects will ripple through several sectors. Semiconductor equipment manufacturers like Applied Materials and ASML should see sustained demand from NVIDIA's need to ramp production. Data center real estate investment trusts (REITs) such as Digital Realty Trust and Equinix will benefit from the construction of new facilities to house this computing power. Companies that provide liquid cooling solutions, critical for high-density AI server racks, are also positioned for growth. The energy sector faces a mixed impact, with increased demand for electricity offset by the push for more efficient computing.

A significant risk to this bullish narrative is the potential for an AI infrastructure oversupply. If enterprise adoption of advanced AI models slows or fails to meet optimistic projections, cloud providers could be left with underutilized, expensive hardware. This would lead to price wars for AI inference services, compressing margins across the sector. Another limitation is the concentration risk for NVIDIA; while this deal is positive, it also makes the company heavily reliant on a small number of hyperscale customers for a large portion of its data center revenue. Market positioning data shows that institutional ownership of NVIDIA remains near all-time highs, but short interest has been creeping up, reflecting these concerns.

Outlook — what to watch next