Key HighlightsAmazon Web Services will receive 1 million GPUs from Nvidia by the conclusion of 2027.Deliveries commence in 2025 and continue through the end of 2027.The agreement encompasses networking equipment, Groq inference processors, and upcoming Blackwell and Rubin architectures.AWS plans to deploy seven different Nvidia chip varieties for AI inference operations.NVDA and AMZN shares both climbed in extended trading after the announcement.The Amazon Web Services agreement represents one of Nvidia’s most substantial single-client semiconductor contracts disclosed to date. A closer examination reveals increasingly compelling details about the partnership. $NVDA × $AMZN — Massive GPU Deal Just DroppedAt GTC 2026, Nvidia and AWS officially announced a deal to deploy 1M+ Nvidia GPUs — including Blackwell & Rubin architectures — across AWS global regions starting this year.Nvidia VP Ian Buck confirmed deliveries run through… pic.twitter.com/kKABFWM1FW— Invest Alpha Pro (@InvestAlphaPro) March 19, 2026In a statement to Reuters, Nvidia Vice President Ian Buck disclosed that GPU deliveries totaling 1 million units will commence in 2025 and extend through 2027. This timeframe aligns precisely with CEO Jensen Huang’s forecast of a $1 trillion addressable market for Nvidia’s Blackwell and Rubin processor families during the identical period.NVIDIA Corporation, NVDAThe partnership extends considerably beyond simple GPU quantities. AWS is acquiring a comprehensive portfolio of Nvidia infrastructure, including Spectrum-X networking components and ConnectX equipment. This development is particularly significant given AWS’s historical reliance on proprietary networking solutions. The integration of Nvidia’s networking technologies into AWS data centers represents a substantial strategic pivot.Amazon Web Services Embraces Comprehensive Nvidia Inference StrategyAI inference — the computational process enabling AI platforms to produce outputs and execute tasks — forms the foundation of this partnership’s technical framework. AWS intends to leverage seven distinct Nvidia chips for managing inference operations.Buck articulated the complexity clearly: “Inference is hard. It’s wickedly hard. To be the best at inference, it is not a one chip pony. We actually use all seven chips.”The Groq processors, unveiled by Nvidia earlier this week following a $17 billion licensing arrangement with an AI semiconductor startup, constitute part of that inference ecosystem. These chips function in tandem with six additional Nvidia processors to provide what the company characterizes as industry-leading inference capabilities.AWS will also receive Nvidia’s Blackwell processors and is anticipated to integrate the forthcoming Rubin platform upon its release. Neither Nvidia nor Amazon has revealed the monetary terms of this arrangement.Both companies’ shares experienced modest gains during after-hours trading Thursday following the disclosure. NVDA declined approximately 1% during regular trading hours, while AMZN dropped about 0.5%.AWS Continues Developing Proprietary Chip SolutionsAmazon maintains its own AI semiconductor development efforts, including the Trainium2 processor. Nevertheless, the company continues partnering with Nvidia for the most resource-intensive computing requirements. These dual strategies appear complementary rather than contradictory.This agreement underscores ongoing substantial capital allocation toward AI infrastructure among leading cloud service providers. AWS is not abandoning its custom silicon initiatives — instead, it’s augmenting them with Nvidia hardware for specialized high-performance scenarios.The Nvidia-AWS partnership was initially announced earlier this week without precise timeline details. Buck’s Thursday remarks to Reuters provided unprecedented clarity: deliveries beginning in 2025, extending through late 2027, encompassing a diverse array of Nvidia offerings spanning compute, networking, and inference capabilities.The post Nvidia (NVDA) Stock Climbs as Amazon AWS Orders 1 Million GPUs Through 2027 appeared first on Blockonomi.