Rise of the Data Center GPU: 5 Predictions for 2019

Why GPUs will become ubiquitous in data center servers across the enterprise.
by Anne Hecht

From AI and HPC to rich media and Windows 10, the ways in which people work and the applications they use are changing dramatically. GPU computing plays a part in all of these areas.

Here are our top five predictions of why GPUs will become ubiquitous in servers across the enterprise.

AI Comes to Virtualized Workflows

Whether it’s generative design, analyzing product quality goals or interactive rendering of images, some form of AI in the workplace is becoming the norm and transforming the way we work. To stay competitive and speed time to market, organizations will need to handle AI and machine learning capabilities with anytime, anywhere access and organizational readiness. All this AI means more data, and more data center GPU use.

Flexible, Application-Centric Data Centers Accelerate Productivity

AI, HPC, VR, photoreal design, Windows 10 — organizations today run a broad range of workloads. New technologies require IT to deliver an agile data center. By repurposing hosts that run VDI during the day to run HPC or other compute workloads at night, organizations can maximize the return on their infrastructure and improve employee productivity. A flexible GPU-accelerated architecture that supports mixed workloads ensures users always get the performance they need while using GPU resources that might have otherwise been idle.

Shift to Hybrid Cloud Brings Innovation

Cloud adoption is growing in the enterprise, and hybrid cloud is leading the way. Growing numbers of companies are pursuing the benefits that hybrid cloud deployments promise — scalability, cost efficiency, agility. Many have come to realize that blending the strengths of public and private clouds requires a combination of software, hardware and systems infrastructure that spans from the cloud to the network edge. AI workloads, graphics-intensive applications or any other from the cloud or data center will require the computing power of virtual GPUs (vGPU) to deliver a seamless experience to employees on virtual workstations, whether they’re running from the on-prem data center, private cloud or public cloud.

Rich Media Drives GPU Demand for Every User

They say video killed the radio star. But has it also killed the print star? Whether for internal training or marketing, videos are processed by the brain 60,000 times faster than text. That’s one reason video is among the most pervasive ways businesses engage with both customers and employees in a meaningful way. Video is just one example of today’s rich media content that people are consuming wherever they go at anytime, from anywhere. For the cloud and VDI deployments, GPU acceleration is critical to ensuring organizations keep up with these multimedia expectations.

Windows 10 Migrations Move Ahead, Even as Windows 7 Remain Popular

As Microsoft continues to push the Windows 7 end-of-life, some companies are reluctant to fix something they don’t see as broken. So, while Redmond would really like businesses to virtualize Windows 10, you can still fire up a Windows 7 Enterprise virtual desktop and receive free Extended Security Updates past the January 2020 end of life for the venerable OS. Microsoft believes this will give customers “more options to support legacy apps while you transition to Windows 10,” but enterprises will still want to prioritize migrations next year as extended support is expensive.

Prepare your organization. Choose the right NVIDIA vGPU solution to deliver the operational, cost and productivity benefits of AI, rich media, hybrid cloud and other growing applications.