PCI Passthrough and SR-IOV on

Overview

Tue, 30 Jun 2026 00:00:00 +0000

Overview PCI passthrough allows virtual machines to access physical PCI devices directly, bypassing device emulation. This approach enables workloads to use the full capabilities of specialized hardware such as GPUs, high-performance network adapters, storage controllers, and other accelerators while maintaining the flexibility of virtualization.

This guide introduces the PCI passthrough architecture in OpenNebula and provides an overview of the supported device types. Detailed configuration procedures are covered in the following guides.

Host Configuration for PCI Passthrough and SR-IOV

Tue, 30 Jun 2026 00:00:00 +0000

This guide describes the common host configuration required to enable PCI passthrough in OpenNebula. These requirements apply independently of the type of PCI device being assigned to virtual machines. Device-specific configuration is covered in the corresponding guides. The guide assumes a Linux host running the OpenNebula KVM hypervisor. While the examples use common Linux utilities and a GRUB-based bootloader, equivalent procedures can be used with other Linux distributions and bootloaders.

Generic PCI Devices

Wed, 02 Sep 2026 00:00:00 +0000

This guide describes how to assign generic PCI devices directly to Virtual Machines. Examples include NVMe controllers, SATA or SAS host bus adapters, USB controllers, audio devices, FPGAs, and other accelerators. For network adapters and GPUs, use the corresponding device-specific guides.

PCI passthrough operates on PCI functions. For example, passing through a storage controller gives the Virtual Machine control of the controller and its attached disks; passing through a USB controller gives it control of the controller and its downstream USB devices. It does not pass an individual non-PCI disk or USB peripheral.

Network Interfaces with PCI Passthrough

Tue, 30 Jun 2026 00:00:00 +0000

This guide describes how to use PCI network devices as OpenNebula network interfaces.

OpenNebula extends the generic PCI passthrough mechanism by integrating PCI network devices with the Virtual Network subsystem. This allows virtual machines to benefit from the performance of direct device assignment while preserving the OpenNebula networking workflow, including IP address management, MAC address allocation, scheduling, and contextualization.

General host preparation, including IOMMU, VFIO, Huge Pages, SR-IOV configuration, and PCI monitoring, is described in the Host Configuration guide.

NVIDIA vGPU & MIG

Mon, 17 Feb 2025 00:00:00 +0000

Configuring the hypervisor for NVIDIA® vGPU and MIG (Multi-Instance GPU) capabilities facilitates centralized management, flexibility in resource allocation, and enhanced security isolation across all virtualized GPU workloads.

BIOS

You need to check that the following settings are enabled in your BIOS configuration:

Enable SR-IOV
Enable IOMMU

Note that the specific menu options where you need to activate these features depend on the motherboard manufacturer.

NVIDIA GPU Passthrough

Thu, 16 Oct 2025 00:00:00 +0000

Overview

This guide describes how to deploy NVIDIA GPUs using PCI passthrough technologies in OpenNebula. OpenNebula supports multiple NVIDIA GPU deployment models, including conventional PCI passthrough, NVIDIA vGPU, MIG-backed vGPU, and NVIDIA Grace platform virtualization. While these deployment models share the same scheduling and deployment workflow, each requires a different Host configuration.

The generic PCI passthrough configuration, including IOMMU, VFIO, Huge Pages, PCI monitoring, and device discovery, is described in the Host Configuration Guide. This guide focuses exclusively on the NVIDIA-specific configuration and deployment procedures.

NVIDIA vGPU and MIG-backed vGPU

Thu, 16 Oct 2025 00:00:00 +0000

NVIDIA vGPU and MIG-backed vGPU

Overview

NVIDIA vGPU enables multiple Virtual Machines to share the resources of a physical GPU by exposing one or more virtual GPU (vGPU) profiles that can be independently assigned to guest operating systems.

Depending on the GPU generation, vGPU profiles can be created using one of two deployment models:

NVIDIA vGPU, where virtual GPU profiles are created directly by the NVIDIA AI Enterprise software stack.
MIG-backed vGPU, where the GPU is first partitioned using NVIDIA Multi-Instance GPU (MIG), and one or more vGPU profiles are created from the resulting GPU Instances.

Regardless of how the profiles are created, OpenNebula automatically discovers the available vGPU profiles during Host monitoring and schedules them as GPU resources that can be assigned to Virtual Machines.

NVIDIA Grace Platform Virtualization (AI)

Thu, 09 Jul 2026 00:00:00 +0000

NVIDIA-Certified Hypervisor

OpenNebula is an NVIDIA-Certified Hypervisor for the Grace Blackwell platform. This certification validates that OpenNebula correctly exposes NVIDIA GPUs and associated PCIe resources to Virtual Machines while preserving the topology, high-performance data paths, and near bare-metal performance required for accelerated AI and compute workloads. For users deploying NVIDIA Grace Blackwell infrastructure, this ensures that OpenNebula has been tested against NVIDIA-defined requirements for running GPU-accelerated workloads in virtualized environments.

Overview

NVIDIA Grace Hopper and Grace Blackwell platforms implement a virtualization architecture that differs from conventional PCI passthrough systems. In addition to assigning GPUs to virtual machines, Grace platforms require a specific PCI topology based on dedicated PCI Root Complexes, SMMUv3 devices, IOMMUFD, and ACPI Generic Initiator NUMA nodes.

NVIDIA NVSwitch (AI)

Thu, 09 Jul 2026 00:00:00 +0000

AMD GPU Passthrough

Tue, 30 Jun 2026 00:00:00 +0000

Work In Progress

GPU passthrough functionality for AMD GPUs is currently under active development. This documentation will be updated once the integration is completed and tested. Please return to this page at a future date to check for updates. Alternatively, if you would like to discuss a demonstration, please contact the OpenNebula sales and customer support team.

Axelera GPU Passthrough

Thu, 16 Oct 2025 00:00:00 +0000

Work In Progress

GPU passthrough functionality for Axelera GPUs is currently under active development. This documentation will be updated once the integration is completed and tested. Please return to this page at a future date to check for updates. Alternatively, if you would like to discuss a demonstration, please contact the OpenNebula sales and customer support team.

OpenNebula NVIDIA Fabric Manager (AI)

Mon, 01 Jan 0001 00:00:00 +0000

Introduction

The OpenNebula NVIDIA® Fabric Manager integration provides a complete solution for managing NVIDIA NVSwitch fabric within a cloud environment.

NVIDIA Shared NVSwitch Virtualization Model

The OpenNebula integration with Fabric Manager integration follows the NVIDIA Shared NVSwitch Virtualization Model. This model uses a Service VM in each hypervisor for managing the NVSwitches. The NVSwitches are added to the Service VM as PCI passthrough devices to configure the selected GPU partitioning. The Guest VMs are configured with PCI passthrough for the GPUs only, without any visibility of the NVSwitches. See the reference architecture in the image below.