Nvidia VMware partnership will extend high-performance graphics throughout the enterprise

Nvidia now has access to VMware’s 76% market share in enterprise computing.

By Jon Peddie

Last week at Nvidia’s annual GPU Tech Conference (GTC) in San Jose, the company announced VMware is adding virtualized GPU (vGPU) support to vSphere, an enhanced suite of tools for cloud computing. That will bring Nvidia GRID technology to its VMware’s Horizon desktop as a service (DaaS) platform.

This is big news many in our industry have overlooked, so it’s worth paying attention to.

VMware CTO Ben Fathi (left) joined Nvidia CEO Jen-Hsun Huang on-stage at the GPU Technology Conference to announce virtualized GPU support. (Source: Jon Peddie Research)

Why is this important?

Not everyone who wants and needs a real workstation can afford one. The installed base of workstations is about 30 million. With GRID and VMware, the potential for adding new users expands by another 30 million designers, engineers, project managers, and power users.

The GPU has become ubiquitous, from the multi?billion dollar gaming industry, to workstations running CAD, and digital media entertainment applications, to high?resolution graphics and video in the palm of your hand. Even mainstream productivity apps like Microsoft Office 2013 demand GPU capabilities, like DirectX compliance. And the emerging influx of rich HTML5 content on the web requires having a capable GPU under the hood of whatever device is doing the browsing.

The reason is simple: a quality interactive visual experience matters, and delivering on that experience requires a GPU. That’s a premise validated throughout the evolution of personal computing. And it is a premise just as valid for any enterprise’s IT plans, whether focused on GPU?equipped clients, like PCs and workstations, or a future with GPUs in the datacenter powering virtual desktops and remote graphics.

To get a virtual GPU to work in a vSphere and DaaS two very critical and unusual things had to happen involving closely guarded IP. VMware had to open up their source to Nvidia, and likewise Nvidia had to open up their GIRD source to VMware

To make DaaS with GRID, Nvidia and VMware embedded Nvidia’s microcode into VMware’s Hypervisor—that’s a really, really, big deal.

Nvidia hardware and software run though the VM stack. (Source: JPR)

VMware’s vSphere has had the capability to pass-through a dedicated GPU to a virtual machine (VM) for a while now. But, with GRID vGPU, any enterprise VMware cloud, private or public, will be able to deploy virtual GPUs into its virtual systems anywhere in the world. VMware users can have a dedicated GPU or a piece of a virtual GPU on any platform the enterprise is supporting with VMware. Needless to say, Nvidia’s available market for GPUs just took a jump. As the company’s CEO and founder Jen-Hsun Huang said: “Partnering with VMware makes it possible for us to virtualize the enterprise end to end.”

Ironically, this is development is so earthshaking and so esoteric, that only a few people get it. However, one day, maybe a year from now, it will be noticed, and folks will say, “Holy cow! When did that happen?

Jon Peddie is president of Jon Peddie Research, which publishes GraphicSpeak.