ReleaseBytes
aws AWS What's New ·

Amazon SageMaker AI inference adds G6 instances in AWS GovCloud

aiawsgovernmentaws-ec2aws-sagemaker
feature announcement

Amazon SageMaker AI inference now supports G6 instances in the AWS GovCloud (US-East) region. Powered by NVIDIA L4 GPUs, these instances offer up to 2x deep learning inference performance compared to G4dn instances, providing government agencies with enhanced capabilities for generative AI workloads. This expansion enables compliance with strict data residency requirements for agencies operating in GovCloud.

  • →G6 instances available on SageMaker AI inference in AWS GovCloud
  • →Improved inference performance for generative AI workloads
  • →Compliance and data residency for government agencies
Features (1) ›
  • G6 instances available on SageMaker AI inference in AWS GovCloud

    Amazon EC2 G6 instances are now available on Amazon SageMaker AI inference in the AWS GovCloud (US-East) region. These instances utilize NVIDIA L4 Tensor Core GPUs and AMD EPYC processors to deliver enhanced deep learning inference performance.

Enhancements (1) ›
  • Improved inference performance for generative AI workloads

    G6 instances offer up to 2x the deep learning inference performance compared to G4dn instances, supporting generative AI workloads like language models, image generation, and computer vision tasks.

Notes (1) ›
  • Compliance and data residency for government agencies

    This expansion allows government agencies and organizations in GovCloud to deploy inference endpoints on G6 instances while adhering to strict compliance and data residency requirements. G6 instances are suitable for production inference workloads that require up to 24 GB of GPU memory.

Read the original announcement →

https://aws.amazon.com/about-aws/whats-new/2026/07/g6-sagemaker-ai-inference/

Related releases