Microsoft Using NVIDIA GRID 2.0 For GPU Accelerated Cloud Computing Content

Posted Sep 30, 2015
49Shares
Share Tweet Submit

NVIDA just announced the availability of their GRID 2.0 GPU compute platform for data-centers a few weeks ago, and now Microsoft is going to make use of that turn-key infrastructure for their Azure cloud platform.

GRID 2.0 to be used as the backbone of Microsoft’s new N-Series of Azure virtual machines.

Azure is Microsoft’s cloud compute platform, something that can be setup to be a companies server and computational assets for anything from hosting an Exchange server to even letting users log-in to virtual desktops to do work, eliminating the need for a local hardware lifecycle.

Though GRID servers have been available, they’ve sort of been a novelty and not necessarily a selling point for many cloud providers. HP, Amazon’s AWS, Google and of course Azure have played down GPU’s available in their servers, instead focusing on other aspects that make cloud computing a good buy compared to running your own hardware.

GRID 2.0 being a major part of Azure sort of changes that, letting you either use it as a backend for mobile app’s that need compute work done that can’t be done locally, or by having it part of a node of servers that can be leveraged for all manner of compute work. Use it to handle encoding of videos, computing Pi to 20 Quadrillion, or anything you could dream of.

This type of design win for NVIDIA is actually very favorable, as it sets up a precedent for even larger companies to use turn-key GPU compute server solutions. Certainly one can price out and build custom servers using any manner of components, but this helps to provide a level of support and compatibility that might be beneficial to enterprise customers.

NVIDIA GeForce GTX 1060 With Pascal GP106 GPU Pictured - Launching in July in 3 GB and 6 GB Variants, Will Compete Against the RX 480

If only there was a free trial of their new GRID 2.0 enabled virtual desktop to test the compute capabilities with some distributed computing projects.

Share Tweet Submit