Introducing Amazon EC2 G3 Instances, the next-generation of GPU-powered instances for graphics-intensive applications

Posted on: Jul 13, 2017

You can now launch G3 instances, the latest generation of Amazon EC2 Accelerated Compute Instances. G3 instances make it easy to procure a powerful combination of GPU, CPU, and host memory for workloads such as 3D rendering, 3D visualizations, graphics-intensive remote workstations, video encoding, and virtual reality applications.

Backed by the NVIDIA Tesla M60 GPUs, G3 instances offer double the CPU power per GPU, and double the host memory per GPU when compared to the most powerful GPU cloud instance available today. This allows you to do complex modeling and 3D visualization analyses such as medical image processing, computer-aided design, or seismic visualization jobs in much less time than possible with any other GPU cloud instance.

The largest G3 instance, g3.16xlarge, features four NVIDIA Tesla M60 GPUs, 64 vCPUs using custom Intel Xeon E5-2686v4 (Broadwell) processors, Amazon Elastic Block Store (EBS) optimized hardware, 488 GiB of host memory, and Enhanced Networking using the Amazon EC2 Elastic Network Adaptor. G3 instances are the first Amazon EC2 instances to support NVIDIA GRID Virtual Workstation capabilities, with streaming support for four monitors each with up to 4K resolution (4096 x 2160 pixels), and hardware encoding to support up to 10 High Efficiency Video Coding (HEVC) H.265 1080p30 streams or up to 18 H.264 1080p30 streams per GPU for faster video frame processing and improved image fidelity.

You can launch G3 instances using the AWS Management Console, AWS Command Line Interface (CLI), AWS SDKs, and third-party libraries. G3 instances are available in three instance sizes: g3.4xlarge, g3.8xlarge, and g3.16xlarge. G3 instances are available in US East (Ohio), US East (N. Virginia), US West (Oregon), US West (N. California), AWS GovCloud (US), and EU (Ireland) with more regions coming soon.

To learn more about G3 instances, visit the AWS Blog.