Overview
This guide provides step-by-step instructions to deploy NVIDIA vGPU software in a Microsoft Windows Server environment. It assumes familiarity with virtualization environments and NVIDIA vGPU technology. For a comprehensive overview of NVIDIA vGPU architecture, licensing, and sizing, refer to the following documentation:
Ensure your deployment uses NVIDIA GPUs compatible with vGPU software and supported on Windows Server. For a complete list of supported NVIDIA GPUs and validated server platforms, please refer to our Microsoft Windows Server release notes.
Some GPUs support display-off and display-enabled modes. For vGPU software deployments, they must be used in display-off mode. For more information, see Switching the Mode of a GPU that Supports Multiple Display Modes.
Ensure your hardware is certified by NVIDIA and compatible with Windows Server. Also, see the hardware requirements for Microsoft Windows Server.
Download valid vGPU licenses.
Refer to the Microsoft Windows Server release notes to find the appropriate vGPU software driver versions to use.
Refer to the Microsoft Windows Server release notes for supported Windows and Linux guest OS versions.
NVIDIA vGPU software supports only 64-bit guest operating systems. No 32-bit guest operating systems are supported.
NVIDIA vGPU software supports assigning multiple vGPUs to a single VM, but the maximum number is limited. For detailed information, see vGPUs that Support Multiple vGPUs Assigned to a VM.
Ensure the following BIOS settings are enabled:
Virtualization support, for example, Intel Virtualization Technology (VT-x) or AMD Virtualization (AMD-V)
SR-IOV
Above 4G Decoding
For Supermicro servers: ASPM Support
For servers that have an AMD CPU:
Alternative Routing ID Interpretation (ARI)
Access Control Service (ACS)
Advanced Error Reporting (AER)
For deployments that will utilize GPU Pass-Through, VT-D/IOMMU is also required to be enabled in the BIOS.