Intel Xe GPU Software

Overview

  • Introduction
  • Architecture
    • Intel® AI Software Stack Architecture
    • Intel® Arc™ Pro B-Series Graphics Architecture
  • Release Notes
    • Intel OMIX Release Notes
      • 0.5.0
      • 0.4.1
      • 0.4.0
      • 0.3.0
      • 0.2.0
      • 0.1.0
    • Container Images Release Notes
      • Intel OMIX Container Release Notes
        • 0.5.0
        • 0.4.1
        • 0.4.0
        • 0.3.0
        • 0.2.0
        • 0.1.0
      • Compute Runtime Container Release Notes
        • 26.35.39758.21
        • 26.31.39395.16
        • 26.31.39395.13
        • 26.22.38646.7
        • 26.18.38308.4
        • 26.14.37833.4
      • PyTorch Container Release Notes
        • 2.15.0
        • 2.14.0
        • 2.13.0
        • 2.11.0
      • vLLM Container Release Notes
        • 0.21.0
  • Support Matrix

Getting Started

  • Docker and vLLM Quick Start Guide
  • Intel OMIX Bare-Metal Installation Guide
  • Kubernetes Setup Guide
  • Red Hat OpenShift Setup Guide
    • Node Feature Discovery
    • Intel GPU Resource Driver (DRA)
    • Intel GPU Base Operator
    • Intel Device Plugins Operator
    • Integration with Red Hat OpenShift AI
      • Configuring Storage for OpenShift AI
      • Installing OpenShift AI Operator
      • Configuring GPU Profiles
      • Notebook Configuration
      • Model Serving with Intel GPUs (vLLM)
      • Summary of All YAML Resources
    • Verifying GPU Access
    • Troubleshooting
  • Tutorials and Examples
  • Container Images
    • Using Intel® OMIX Container Images

AI Frameworks and Workloads

  • Running Inference Workloads
    • Running vLLM Inference Workloads in Docker Containers
    • Running SGLang Inference Workloads in Docker Containers

Configuration and Optimization

  • Performance and Optimization
    • Profiling and Instrumentation
      • PyTorch Profiler
        • PyTorch Profiling
        • Hugging Face Profiling
        • vLLM Profiling
        • SGLang Profiling
        • Distributed Collective Traces
        • Scope Profiler and Hardware Metrics
        • Troubleshooting
      • Intel® VTune™ Profiler
  • Virtualization with SR-IOV
    • Preparing the System for SR-IOV
    • Enabling SR-IOV Virtual Functions
    • Configuring VRAM Allocation for Virtual Functions
    • Installing the Virtualization Stack and Preparing Cloud-init Configuration
    • Creating and Configuring Virtual Machines with GPU Virtual Functions
    • Reusing a Prepared Virtual Machine Image
    • Managing GPU Virtual Machines

Legacy

  • LTS Release Notes
    • 2523.104
    • 2523.78
    • 2523.59
    • 2523.41
    • 2523.40
    • 2523.31
    • 2350.150
    • Provided Packages
    • Delivered Projects
  • Intel Max and Flex Series Support Matrix
  • Supported Hardware
    • Xe KMD Driver GPUs
    • i915 KMD Driver GPUs
    • Legacy GPUs
  • Installation Guides
    • Installing Drivers on Intel Max and Flex Series
      • Installing LTS 2523.x
      • Installing LTS 2350.x
      • Verifying Installation
      • Configuring Kernel Boot Parameters
      • Configuring Secure Boot
    • Installing Packages from Ubuntu PPA
  • Setup Guides
    • Configuring GPU Debugging Tools
    • Configuring the Isolated Desktop Solution
  • Troubleshooting and FAQ
    • Frequently Asked Questions (FAQ)
    • Troubleshooting
Intel Xe GPU Software

Container Images Release Notes

Review release notes for the following container images:

  • Intel® Open Middleware Xe (Intel® OMIX)

  • Intel® Graphics Compute Runtime for oneAPI Level Zero and OpenCL™ Driver

  • PyTorch* on Intel® Hardware

  • vLLM


On this page