The G7e instances provide up to 2.3 times the inference performance of the prior G6e generation, utilizing up to eight NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs with 96GB memory each. Coupled with 5th Generation Intel Xeon processors and up to 192 vCPUs, they support demanding AI model deployment including large language, multimodal generative, and agentic AI models.
G7e instances also feature enhanced networking bandwidth up to 1600 Gbps and NVIDIA GPUDirect technologies for improved multi-GPU and multi-node performance. These capabilities cater to spatial computing and workloads requiring simultaneous graphics and AI processing, expanding availability to key global AWS regions.
Source & further reading
Amazon EC2 G7e instances now available in Europe (London) region ↗Source published 2026-05-07. 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.