Virtualization is a key tool to address this challenge, offering a great user experience for employees logging in from home, while ensuring sensitive data is kept safe.
While virtual desktop infrastructure was once considered a low-quality option, the virtualization of NVIDIA GPUs has made remote workstation and server-class performance a reality for even the most compute-intensive workloads.
NVIDIA virtual GPU (vGPU) software enables IT teams to help remote employees stay productive, even when they’re working on sophisticated graphics applications or AI and machine learning workloads. Companies supporting remote workers can learn more about how we’re expanding our free vGPU licenses to 500 users for 90 days.
What Is a vGPU?
Traditional physical computing devices like a PC, laptop or server use GPUs to perform complex workflows, such as running 3D applications or streaming videos. However, this isn’t ideal for professionals who need to work remotely or while on the go because the GPUs are tied to the physical devices, limiting mobility and flexibility.
Installed on a physical GPU in a server, NVIDIA vGPU software creates virtual GPUs that can be shared and allocated between multiple virtual machines. This allows IT to create software-defined GPU acceleration for any workflow, on any device and any location.
Virtualizing a data center with GPUs improves performance for graphics-intensive applications and workflows, allowing companies to use VDI to cost-effectively scale performance to all their employees.
In addition, virtualization allows IT teams to optimize their infrastructure so they can meet the unique needs of their business. With vGPUs, IT can either share GPU performance across multiple virtual machines, or power demanding workloads by assigning multiple GPUs to a single machine.
While companies are now supporting more remote workers than ever before, few have the same requirements for supporting their staff working from home. Different NVIDIA vGPU software licenses are available to meet the different needs of multiple virtualization use cases.
Virtual GPUs for the Knowledge Worker: NVIDIA GRID vPC, GRID Virtual Apps
Most people use multiple applications, high-resolution monitors, productivity tools and browser windows on a daily basis. Responsive performance is critical to a good user experience. GPU acceleration is essential to powering these workflows, especially in virtualized environments.
NVIDIA GRID virtual PC (GRID vPC) and virtual apps (GRID vApps) use the power of GPUs to deliver immersive, virtualized Windows 10 workspaces to every remote user, providing maximum optimization for office apps and multimedia.
Providing performance as good as a physical PC, NVIDIA GRID cost-effectively delivers greater security, improved scalability and reliable accessibility on any device, including tablets. This allows professionals to be more productive using modern applications because they can work in any way they want, wherever they go.
A Platform for Creative and Technical Users: Quadro Virtual Data Center Workstation
Whether creating stunning visual effects or loading a CAD file containing every component of a building or vehicle, professionals need a powerful visual computing platform that keeps pace with their workflows.
NVIDIA Quadro Virtual Data Center Workstation (Quadro vDWS) delivers workstation-class performance for professional graphics applications, so people can stay productive while on the go or working from home offices.
Quadro vDWS provides GPU-accelerated performance from the data center to give creative and technical professionals greater productivity, flexibility and mobility. This also helps geographically dispersed teams to collaborate and securely access apps and data in real time, from anywhere and on any connected device.
And with support for up to two 8K displays, architects, engineers and designers can view their designs with more details in context without the need to zoom in and out of images.
Virtualization for AI, Deep Learning and Data Science: NVIDIA Virtual Compute Server
From AI and deep learning to data science and high performance computing, compute-intensive workloads require complex technology stacks that are often challenging to deploy and maintain.
With NVIDIA Virtual Compute Server (vComputeServer), users gain the GPU-accelerated performance they need to run these workloads, along with the security and management benefits of virtualization.
With support for all major hypervisor virtualization platforms, including VMware, Red Hat, Nutanix and Citrix, vComputeServer makes it easy for admins to streamline management and migration by using the same tools for their GPU clusters as they do for the rest of their data center, helping lower overall operational costs.
IT departments can easily allocate compute resources with GPU sharing and aggregation. And with GPU sharing, multiple virtual machines can be powered by a single GPU to maximize utilization and affordability.