Yesterday, NVIDIA launched a new virtualization software, dubbed Quadro Data Center Workstation (vDCW).
In its announcement, the GPU maker states that the Quadro vDCW can “turn NVIDIA Tesla GPU-accelerated servers into powerful workstations.”
Most NVIDIA users are familiar with Quadro as the company’s workstation-targeted GPU line, and TESLA as its server-class GPU line. How are these two distinctly different types of GPUs coming together in the new software? In the land of VDI (virtual desktop infrastructure), where Tesla servers can mimic the behaviors of Quadro-powered workstations.
Quadro in Character, Tesla Under the hood
To host a GPU-accelerated VDI, you need Tesla-powered server hardware. To get the benefits of a professional-class NVIDIA GPU-powered workstation, you need a Quadro-powered machine. With NVIDIA’s Quadro vDCW software, you can now spin up virtual machines (VMs) that behave like Quadro-powered workstations but are sitting on Tesla server hardware.
Creating a Quadro-like experience in the VM has certain benefits, according to NVIDIA. “With Quadro vDWS on Tesla-powered servers, businesses can tackle larger datasets, power the most demanding applications and meet the need for greater mobility,” said Bob Pette, Vice President of Professional Visualization at NVIDIA.
Quadro-powered workstations are certified to work with recognized software packages like Siemens NX and Dassault Systèmes SOLIDWORKS. With vDCW, IT managers can give remote workers the same experience by delivering VMs that work and behave like Quadro-powered workstations. But users may access them through lightweight mobile devices or consumer-class PCs.
The NVIDIA vDCW supports a comprehensive line of Tesla GPUs, including Tesla P4, P40, P100, and the new P6. It’s also tuned for CUDA-based workloads and applications.
Grid for Higher Density VDI
The GPU maker is also updating its Grid VDI software with support for GPUs that offer higher density — “50% more vGPUs per physical GPU,” according to its press briefing materials.
“The Tesla P40 supports up to 24 instances of 1GB virtual desktop profiles, which is a 50% increase when compared to the Tesla M60,” according to NVIDIA.
The company introduces a new set of monitoring tools to provide vGPU visibility at the host or guest level. “Application-level monitoring capabilities allow IT to intelligently design, manage, and support their end users’ experience. New integrations with VMware vRealize Operations (vROps), Citrix Director and XenCenter enable flexibility and control from a single, unified view,” writes NVIDIA.
The monitoring helps IT managers identify which users need more graphics and compute horsepower, thus enabling them to reconfigure VMs or redistribute available resources to match the users’ workload.
For more on Quadro vDCW, visit its dedicated page here.