The general availability of G6 instances in new regions includes Asia Pacific (Tokyo, Mumbai, Sydney) and Europe (London, Paris, Frankfurt, Stockholm, Zurich). These instances incorporate up to eight NVIDIA L4 Tensor Core GPUs and 24 GB GPU memory along with third-generation AMD EPYC CPUs.
G6 instances deliver approximately double the deep learning inference performance of prior G4dn instances. They serve use cases like generative AI fine-tuning, natural language processing, computer vision, and recommendation systems, allowing interactive model development and deployment workflows within SageMaker notebooks and Studio environments.
Source & further reading
Announcing Region Expansion of G6 instances on SageMaker Notebook Instances ↗Source published 2026-05-12. 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.