Announcing general availability of Amazon EC2 G6f instances with fractional GPUs
Today we are announcing the general availability of Amazon EC2 G6f instances, our first GPU instances provisioned with GPU partitioning powered by NVIDIA L4 Tensor Core GPUs. G6f instances can be used for a wide range of graphics workloads. G6f instances offer GPU partitions as small as one-eighth of a GPU with 3 GB of GPU memory giving customers the flexibility to right size their instances and drive significant cost savings compared to EC2 G6 instances with a single GPU.
Customers can use G6f instances to provision remote workstations for Media & Entertainment, Computer-Aided Engineering, and for ML research, and game streaming. G6f instances are available in 5 instance sizes with half, quarter, and one-eighth of a GPU per instance size, paired with third generation AMD EPYC processors offering up to 12 GB of GPU memory and 16 vCPUs.
Amazon EC2 G6f instances are available today in the AWS US East (N. Virginia and Ohio), US West (Oregon), Europe (Stockholm, Frankfurt, and London), Asia Pacific (Mumbai, Tokyo, and Sydney), Canada (Central), and South America (Sao Paulo) regions. Customers can purchase G6f instances as On-Demand Instances, Spot Instances, or as a part of Savings Plans.
To get started, visit the AWS Management Console, AWS Command Line Interface (CLI), and AWS SDKs, and launch G6f instances with NVIDIA GRID driver 18.4 or later. Additionally, you can connect to your G6f instances seamlessly using Amazon DCV, enabling remote desktop access from anywhere. For Amazon DCV, please refer to the Amazon DCV documentation. To learn more, visit the G6 instance page.