Nvidia GRID GPU Added To AWS Cloud Service

By - Source: Toms IT Pro

Nvidia and Amazon Web Services (AWS) announced the addition of Nvidia's GRID GPU technology to AWS's cloud infrastructure as a service. Dubbed theG2 Instance type, AWS announced the availability of the Nvidia GRID (GK104 "Kepler") Graphics Processing Unit (GPU) to the AWS line of instances.

According to Nvidia, the Nvidia GRID GPU was designed specifically for cloud computing and AWS agrees. "AWS sees a growing benefit for adding GPUs to our cloud," said Matt Wood, general manager of Data Science at AWS. "The NVIDIA GRID GPUs in our new G2 instances enable graphical applications to be rendered in the AWS cloud and streamed to a world with increasing internet bandwidth and proliferation of device types."

The new G2 instance is in addition to AWS current CG1 instance which, according to the announcement on AWS blog, "is a great fit for HPC (High Performance Computing) workloads."  The CG1 instance was introduced in 2010 primarily to fill a scientific rather than a gaming/streaming video need.

The GRID GPUs include a H.264 encoding engine that performs real time compression for high resolution graphics that can be streamed to any internet-connected display including smartphones, PCs and smart TVs. Programs that currently require high end workstations for tasks such as CAD, graphic design, animation, 3D modeling, and other engineering and scientific applications can be run on any device that supports an internet connection.

AWS has been working with various vendors in preparation for this debut and a few applications are listed on the AWS blog which include:

  • Autodesk Inventor (CAD engineering),
  • Revit (Building and Construction Design),
  • Maya (3D Animation),
  • 3Ds Max (3D Modeling and Design),
  • Playcast Media (cloud gaming service),
  • Agawi True Cloud (games and mobile apps),
  • Calgary Scientific Resolution MD (visualization of medical imaging data).

Jeff Barrs, the author of the AWS blog, concluded the announcement with an anticipative request, "I am really looking forward to seeing some cool new applications (and perhaps even entirely new classes of applications) running on these instances. Build something cool and let me know about it!"

About the Author

Bill Oliver has been working in Healthcare for the past 30+ years in a variety of management roles including Material Management, Purchasing, Nurse Registry, and IT. In the past 12 years his focus has been on the business end of IT Contracts, Software Licensing and Purchasing.

More by Bill Oliver

While this platform definitely has potential, it appears to be a platform in need of developers willing to build on it or move their applications to it. As an incentive to software developers who want to make the jump, Nvidia also indicated in their announcement, "To make it easy for software companies to deploy applications onto G2 instances, Los Angeles-based cloud rendering company, OTOY has enabled a Windows- and Linux-based Amazon Machine Image (AMI) with OTOY's ORBX middleware. SaaS companies can install their applications into OTOY's AMI and begin streaming to web browsers within minutes."

The G2 instances are currently available in the US East (N. Virginia), US West (Northern California and Oregon) and EU (Ireland) regions. The instances can be launched as on-demand or spot instances and can also be purchased as reserved instances.  The cost for each instance is based on region, OS, and other criteria and that pricing can be found here.