Nvidia Launches New Chip for Cloud VDI Systems

By Chris Preimesberger  |  Posted 2012-05-16

Pretty much everything else in IT already is virtualized, so why not a processor?

Nvidia on May 15 introduced what it called the "world's first virtualized GPU," and aimed it squarely at accelerating graphics input/output for virtual desktop deployments in cloud systems.

The announcement was made during the company's GeForce customer and partner conference a couple of days ahead of the company's annual stockholders' meeting, set for May 17 in San Jose, Calif.

Virtual desktop systems have been around for more than a dozen years. They work well in some use cases, certainly, but in many of them, performance has been lackluster and the systems themselves are difficult to install and maintain.

All these issues would go away in a flash if the processors were fast enough to handle all the traffic from the end-user devices into the data center and back. The main problem all this time has been that PC and server processors have not been powerful enough--or optimized for cloud deployments. Nvidia has been tackling this for several years, and that work has resulted in its new VGX processor.

Using Nvidia's new VGX card-based platform in the data center, enterprise employees can access a cloud system from any device--thin client, laptop, tablet or smartphone--regardless of its operating system.

Nvidia, long known for its high-end graphics processors used in video game machines such as the Xbox and in laptops--from Hewlett-Packard, Dell and Acer, to name three--claims that VGX offers a responsive experience for the full spectrum of applications previously only available on an office PC.

"Nvidia VGX truly does represent a new era in desktop virtualization," Jeff Brown, general manager of the Professional Solutions Group at Nvidia, told eWEEK. "It delivers an experience nearly indistinguishable from a full desktop."

TheVGX is based on three key home-developed innovations, according to Nvidia:

  • VGX Boards: These are designed for hosting large numbers of users in an energy-efficient way. The first Nvidia VGX board is configured with four GPUs and 16GB of memory, and fits into the industry-standard PCI Express interface in servers.
  • VGX GPU Hypervisor: This software layer integrates into commercial hypervisors, such as the Citrix XenServer, enabling virtualization of the GPU.
  • User Selectable Machines (USMs): This manageability option allows enterprises to configure the graphics capabilities delivered to individual users in the network, based on their demands. Capabilities range from true PC experiences available with the Nvidia standard USM to enhanced professional 3D design and engineering experiences with Nvidia Quadro or Nvidia NVS GPUs.

The VGX enables up to 100 users to be served from a single server powered by one VGX board, improving user density on a single server, compared with traditional virtual desktop infrastructure (VDI) solutions, Brown said. It also reduces such issues as latency, sluggish interaction and limited application support, all of which are associated with traditional VDI solutions, Brown said.

Chris Preimesberger is eWEEK's Editor for Features and Analysis. Twitter: @editingwhiz

Rocket Fuel