The company claims the new instances offer double the CPU power per GPU, and double the host memory per GPU, than the most powerful GPU cloud instance available today.
Customers of Amazon Web Services will now be able to use its next generation of GPU-powered Amazon Elastic Compute Cloud instances.
The G3 Instances for Amazon EC2 are designed to make it easy to procure CPU and RAM for workloads such as 3D rendering, 3D visualisations, graphics-intensive remote workstations, video encoding, and virtual reality applications, according to the company.
The new instances, which are backed by NVIDIA Tesla M60 GPUs, are said to offer double the CPU power per GPU, and double the host memory per GPU, than the most powerful GPU cloud instance available today.
“AWS was the first to offer cloud instances customized for graphics-intensive applications. In 2010, we launched the CG1 instance type to provide a cost-effective, high-performance instance for graphics-heavy applications, and today, G3, our third generation GPU-powered instance, serves the most demanding graphics workloads such as 3D rendering and data visualization,” said Matt Garman, Vice President, Amazon EC2.
“Today, AWS provides the broadest range of cloud instance types to support a wide variety of workloads. Customers have told us that having the ability to choose the right instance for the right workload enables them to operate more efficiently and go to market faster, which is why we continue to innovate to better support any workload.”
AWS says that customers requiring things like graphics-intensive applications require the kind of performance that G3 can offer, with the company saying that the largest G3 instance offers twice the CPU power and eight times the host memory of the previous generation graphics instance (G2).
The g3.16xlarge will offer four NVIDIA Tesla M60 GPUs with 64 CPUs using custom Intel Xeon E5-2686v4 (Broadwell) processors, 488 GB of RAM, and Enhanced Networking using the Amazon EC2 Elastic Network Adaptor. The offering also features support for NVIDIA GRID Virtual Workstation capabilities, and hardware encoding to support up to 10 H.265 (HEVC) 1080p 30 streams per GPU.
G3 instances can be launched through the AWS Management Console, AWS Command Line Interface, AWS SDK’s, and third-party libraries. Three sizes are available: g3.4xlarge, g3.8xlarge, and g3.16xlarge.
Users in regions US East (Ohio), US East (N. Virginia), US West (Oregon), US West (N. California), AWS GovCloud (US), and EU (Ireland) will have first access to the new instances, with more to be added in the coming months.