The G7e EC2 instances feature up to eight RTX PRO 6000 Blackwell GPUs with 96 GB memory each, paired with 5th Gen Intel Xeon processors, delivering up to 2.3 times the inference performance of previous generation G6e instances. They support high-performance workloads including large language models, agentic AI, multimodal generative AI, and spatial computing requiring both graphics and AI acceleration.
Additional capabilities include NVIDIA GPUDirect Peer to Peer for improved multi-GPU workloads and GPUDirect RDMA reducing latency in multi-node clusters. Availability spans Tokyo and several U.S. AWS regions, catering to customers needing powerful AI and graphics processing in the cloud.
Source & further reading
Amazon EC2 G7e instances now available in Asia Pacific (Tokyo) region ↗Source published 2026-02-19. This report was prepared with AI assistance from the linked primary-source material. Claims and limitations should be checked against the original announcement or research.