AWS and NVIDIA expand partnership to promote innovation in AI

NVIDIA and Amazon Web Services (AWS) will soon offer the NVIDIA Blackwell GPU platform, enhancing AI capabilities and providing advanced infrastructure, software, and services.

Today, NVIDIA and Amazon Web Services (AWS), a division of Amazon.com, announced that AWS will soon provide the recently introduced NVIDIA Blackwell GPU platform. In order to help customers unlock new generative artificial intelligence (AI) capabilities, AWS will now offer the NVIDIA GB200 Grace Blackwell Superchip and B100 Tensor Core GPUs, expanding their long-standing strategic collaboration. Together, the companies will deliver the most advanced and secure infrastructure, software, and services.

Elastic Fabric Adapter (EFA) petabit scale networking, Amazon Elastic Compute Cloud (Amazon EC2) UltraCluster hyper-scale clustering, NVIDIA's newest multi-node systems with the next-generation NVIDIA Blackwell platform and AI software, and AWS's Nitro System and AWS Key Management Service (AWS KMS) advanced security are just a few examples of how NVIDIA and AWS continue to combine the best of their technologies. The combination of these technologies offers customers the necessary infrastructure and tools to design and execute real-time inference on large language models more efficiently, massively, and affordably than with previous-generation NVIDIA GPUs on Amazon EC2.

AWS will offer the NVIDIA Blackwell platform, featuring the GB200 NVL72 and 72 GPUs and 36 Grace CPUs. This platform allows customers to scale to thousands of GB200 Superchips through Amazon's robust networking, enhanced virtualization, and hyper-scale clustering. This will enable faster completion of inference workloads for language models with multi-trillion parameter demands.