Key Highlights
- AWS will acquire 2 million more Nvidia GPUs scheduled for 2027-2028 rollout
- This expands Amazon’s previous March commitment of 1 million Nvidia chips
- GPU models include Nvidia Blackwell Ultra, Rubin, and Rubin Ultra processors
- Amazon will implement Nvidia’s complete physical AI infrastructure for warehouse robotics
- After-hours trading saw Nvidia shares climb 4%, while Amazon advanced 0.7%
Amazon Web Services plans to integrate 2 million more Nvidia GPUs throughout its cloud infrastructure between now and 2028, according to a joint announcement made Wednesday by both corporations. The revelation sent Nvidia shares surging 4% during extended trading sessions, with Amazon stock posting a 0.7% increase.
These processors are scheduled for integration during 2027 and 2028. The expansion builds upon Amazon’s March announcement committing to 1 million Nvidia GPUs, which are currently being deployed across AWS facilities throughout this year.
This expanded agreement encompasses Nvidia’s cutting-edge Blackwell Ultra, Rubin, and Rubin Ultra GPU architectures. These premium processors are engineered specifically for AI model training and inference operations, with individual units carrying retail pricing reaching tens of thousands of dollarsāthough enterprise-scale purchasers generally secure volume-based pricing arrangements.
Neither Amazon nor Nvidia disclosed specific financial details regarding the transaction.
During Nvidia’s quarterly earnings presentation Wednesday afternoon, CFO Colette Kress validated the partnership extension. According to Kress, Amazon’s cloud division “is deploying an additional two million GPUs starting this quarter through the second of FY29.”
Amazon Embraces Nvidia’s Physical AI Platform for Robotics
Kress additionally disclosed that Amazon intends to implement Nvidia’s comprehensive physical AI technology stack throughout its warehouse automation systems. This development introduces an additional dimension to their collaboration beyond traditional data center computing applications.
Nvidia’s CEO Jensen Huang discussed the expanded alliance during a Wednesday evening CNBC interview. “This is going to be a great partnership,” Huang emphasized.
Meanwhile, Amazon continues developing proprietary semiconductor solutions via Annapurna Labs, producing both custom CPUs and Trainium, an AI acceleration chip positioned as an alternative to Nvidia’s GPU offerings. Industry reports indicate the upcoming Trainium generation will incorporate Nvidia networking infrastructure.
Hyperscale Cloud Providers Maintain Nvidia Dependence
Despite Microsoft and Alphabet’s Google pursuing internal chip development programs, both companies continue as significant Nvidia clients. Amazon follows this identical trajectory.
The major cloud infrastructure providers are wagering that proprietary silicon solutions can optimize performance for their particular operational requirements. Nevertheless, Nvidia’s GPU technology continues dominating as the preferred hardware platform for enterprise-scale AI model creation.
Combining Wednesday’s announcement with the March agreement, Amazon’s aggregate Nvidia GPU procurement now totals 3 million processors.
The timing of Wednesday’s revelation coincided with Nvidia’s quarterly financial reporting. The 4% after-hours stock appreciation demonstrates investor sensitivity to major cloud provider partnership developments.
Jensen Huang’s post-earnings CNBC interview highlighted the strategic significance of the Amazon collaboration, with leadership from both organizations publicly endorsing the amplified partnership scope.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
