What are vGRID GPU Desktops?

The vGRID Desktop offers a true workstation replacement option with workstation class 3D graphics using virtual graphics acceleration (vGPU) enabled by NVIDIA® Tesla GPU technology to support high-performance graphics and compute workloads.

Utilising the latest Graphics Processing Unit (GPU) technologies from NVIDIA, the vGRID Desktop service delivers workstation-like performance with all the benefits of centralised cloud services.  This means you can enjoy the benefits of on-demand deployment, daily backups and remote access from anywhere using PC over IP (PCoIP) technologies.

How the technology works

Delivering high performance workstation capabilities from a cloud platform requires specialist hypervisor support and industry leading hardware technologies.  vGRID Desktop makes use of NVIDIA Pascal GPU (Tesla P6) hardware installed in each of the VMware Horizon View hosts (clustered). The Pascal GPU is designed specifically for virtualizing desktop graphics.

The use of the proven NVIDIA Graphics Driver in the virtual machine ensures application compatibility with traditional non-virtualised workstation environments and the availability of NVIDIA supported OpenGL and DirectX versions.

NVIDIA GRID vGPUVMware Horizon View

vGRID Desktops use VMware Horizon View to provide an exceptional end user experience. Bandwidth consumption will be lowered through the use of multimedia redirection of H.264 advanced video coding for high-definition content, and the efficiency improvements to the Horizon View client for Linux, offers users an enhanced experience.

Improvements to remote desktops delivered through HTML5 capable web browsers will offer higher quality video playback, faster typing response, and higher scalability.

Additionally, support is available for the latest end-user technologies, including Windows 10, Apple iOS 11 and Google Chromebooks.

NVIDIA vGPU Architecture

The high-level architecture of NVIDIA vGPU is illustrated in below. Under the control of the NVIDIA Virtual GPU Manager running under the hypervisor, physical GPUs are capable of supporting multiple virtual GPU devices (vGPUs) that can be assigned directly to guest VMs.

NVIDIA vGPU ArchitectureGuest VMs use NVIDIA vGPUs in the same manner as a physical GPU that has been passed through by the hypervisor: an NVIDIA driver loaded in the guest VM provides direct access to the GPU for performance-critical fast paths, and a paravirtualised interface to the NVIDIA Virtual GPU Manager is used for non-performant management operations.

Compatability

NVIDIA GRID vGPU technology allows multiple virtual machines to share the power of a single GPU to deliver rich graphics experiences with professional 2D and 3D applications.  Many global companies already leverage this capability in their IT solutions, Among them: European aircraft-manufacturing giant Airbus, healthcare provider MetroHealth, and global construction and engineering firm CH2MHILL.

Support for various 2D and 3D applications is broad and continues to grow, below is a list of some commonly virtualised applications:

AECOsim Building Designer, Allplan, ANSYS Fluent, Autodesk 3ds Max, Autodesk AutoCAD, Autodesk Revit, Bentley MicroStation, ESRI ArcGIS, NEMETSCHEK, SketchUp, MitTek Sapphire, VertexBD, Autodesk ACA, Revit, Chief Architec, Unity.

ESRI Arc GIS

MiTek Sapphire