The State of the GPU-Powered Workspace in 2017

This is the first of a two-part series discussing the major NVIDIA GRID milestones of 2016, and some predictions for the GPU-powered workspace in 2017.
by Tony Paikeday

As we near the end of the year, it’s a great time to think about what the year ahead brings, and what it means to customers, partners and our ecosystem driving this GPU-enabled revolution in end-user computing. In this two part series, we’ll explore both – a recap of this year’s most important milestones, and some predictions of what to look for in 2017.

2016 Was About Enterprise-Grade & Cloud-Ready

We end 2016 witnessing the rise of the GPU as the most important resource supporting the hottest technology trends, including deep learning, AI, VR, autonomous vehicles and virtual workspaces. But before we look forward, it’s also a good time to review how we got here.

This year marked a number of ground-breaking milestones that allowed us to deliver a platform our customers recognize as enterprise-grade and cloud-ready. That means offering the performance and manageability that a solution needs before it can be entrusted with mission-critical workload at scale in any cloud, private or public. Let’s look at the building blocks that made this happen:

NVENC for VMware Horizon and Citrix XenDesktop = Doing More with Less

Offloading protocol encoding from CPU to GPU really means doing more with less. More VDI host scale (users) with less CPU. More responsiveness with less protocol latency. More performance with less network bandwidth consumed. It’s also about great technology partnerships with the industry’s leaders to deliver a 1+1=3 for performance, scale and remote user experience.

GPU Monitoring = Better UX with Lower TCO

IT is always finding itself in the cross-hairs when it comes to resolving VDI user experience issues, as they walk the tightrope of delivering acceptable performance without running up cap-ex. When we added the industry’s first vGPU monitoring, we gave IT the ability to use data-driven analytics to ensure they could right-size their infrastructure design from the get-go, for optimal user experience, lowest cost and faster remediation of UX issues.

Tesla M10 = vGPU Economics and Performance for Every User, Any App

While GRID historically played in the trusted space of enabling secure workflow, productivity and mobility for graphics workstation users, vGPU scale was not a primary concern. But as organizations started to ponder how to extend the reach of virtual desktops and apps, the need to match VDI host density with corresponding vGPU scale became table stakes. And so it was the M10 was born, delivering on the need to provide cost-effective, GPU-accelerated performance for not just power users, but every user, and every app, from the mission-critical to the mundane.

A GPU in Every Cloud = GPU Workspaces with Flexibility, Choice and Speed

As we close out the year, we’ve started to unfold our vision of the GPU-accelerated cloud. This means taking this enterprise-grade platform, and making it pervasively accessible to any and every cloud from which our customers choose to consume virtual desktops and apps. When we announced seven partners delivering on that vision, along with very recent announcements from Google, AWS and Azure, we gave notice to the industry that the GPU-Cloud Era is here, and it’s a table stakes underpinning for any cloud-hosted desktop or app being offered in 2016 and beyond.

These four pillars gave us a working foundation in 2016, upon which we’re building the future. So what do we think 2017 has in store for end-user computing? We’ll take a look forward in part two of this series.