Delivered Projects

Intel® software for general-purpose GPU capabilities provides packages for supported operating systems to ensure full functionality of Intel GPUs. These packages come from various projects responsible for supporting drivers and runtime environments, managing graphics computations and compilations, accelerating media processing, managing system and memory, and monitoring performance.

Some packages are developed entirely by Intel, while others are maintained by open-source communities. Since not all users require every component, installation guides specify the essential packages for most users, while additional packages remain optional to support specific features. All projects and packages are regularly updated and integrated into new releases.

To check which packages belong to each project, refer to Provided Packages. If you encounter issues with a specific package, you can report them in the relevant project’s repository.

The following projects contribute to the Intel® software for general-purpose GPU capabilities:

  • Intel® GPU Firmware

    A part of a collection of kernel mode drivers enabling support for Intel GPUs.

  • Intel® Graphics Driver Backports for Linux* OS (i915)

    A backport project that offers early access to Intel GPU features not integrated into the mainline Linux kernel. It supports Intel data center GPUs and Intel® Arc™ A-Series Graphics (Alchemist). For Alchemist GPUs, the backported i915 driver is focused on enabling specific capabilities, such as GPU debugging, while excluding others, such as display support. For more information, see Kernel Driver Types.

  • Intel® oneAPI Level Zero

    A project that offers direct-to-metal interfaces for offloading workloads to accelerator devices. Its programming interface is highly adaptable, allowing customization for various device requirements. It supports an extended set of language features, including function pointers, virtual functions, unified memory, and I/O capabilities, enabling broader compatibility and flexibility.

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

    A project supporting computations for Intel graphics hardware architectures using oneAPI Level Zero and Open Computing Language (OpenCL) APIs.

  • Intel® Graphics Compiler for OpenCL™

    An LLVM-based compiler for OpenCL targeting Intel graphics hardware architecture.

  • Intel® C for Metal Compiler

    A project containing source code for Low-Level Virtual Machine (LLVM), a toolkit for the construction of highly optimized compilers, optimizers, and run-time environments. It contains tools, libraries, and header files needed to process intermediate representations and converts them into object files.

  • Intel® Video Processing Library

    A project that provides access to hardware-accelerated video decode, encode, and processing capabilities on Intel GPUs to support AI visual inference, media delivery, cloud gaming, and virtual desktop infrastructure use cases.

  • Intel® Video Processing Library GPU Runtime

    A runtime implementation of Intel® Video Processing Library for Intel GPUs. It provides access to hardware-accelerated video decode, encode and filtering.

  • Intel® Video Processing Library Tools

    Tools that provide command-line access to hardware-accelerated video decoding, encoding, and processing capabilities on Intel GPUs.

  • Intel® Media Driver for VAAPI

    A user-mode driver for the Video Acceleration API (VAAPI) that enables hardware-accelerated decoding, encoding, and video post-processing on GEN-based Intel graphics hardware.

  • Mesa 3D Graphics Library

    An open-source graphics library that provides 3D rendering capabilities. It supports a variety of graphics APIs, including OpenGL for Embedded Systems (OpenGL ES), Video Decode and Presentation API for Unix (VDPAU), and Vulkan. A variety of device drivers allows using the Mesa libraries in many different environments ranging from software emulation to complete hardware acceleration for modern GPUs.

  • Intel® XPU Manager and XPU System Management Interface

    A software tool for monitoring and managing Intel data center GPUs, designed to simplify administration, maximize reliability and uptime, and improve usage. It provides insights into hardware health, performance, and power usage, aiding system administrators in optimizing and maintaining Intel-based systems.

  • Video Acceleration API

    An implementation for Video Acceleration API providing access to graphics hardware acceleration capabilities for video processing. It allows streamlining video operations across various graphics hardware.

  • Video Acceleration API Utilities Library

    A set of utilities and tools for the Video Acceleration API, including unit tests based on the Google Test Framework.

  • Intel® Graphics Memory Management Library

    A project that provides device-specific buffer management for Intel® Graphics Compute Runtime for oneAPI Level Zero and OpenCL™ Driver and Intel® Media Driver for VAAPI.

  • Intel® Graphics System Controller Firmware Update Library

    A library that provides an API for updating firmware on Intel discrete graphics devices. It uses the cross-platform ME TEE library to communicate with the Graphics System Controller (GSC) device.

  • Intel® Metrics Discovery Application Programming Interface

    A user-mode library that provides access to detailed GPU performance data. It offers an OS abstraction layer that ensures interoperability across different environments.

  • Intel® Metrics Library for Metrics Discovery API

    A user-mode driver helper library that provides access to GPU performance counters.

  • Intel® ME TEE Library

    A library designed to access the firmware of Converged Security Engine (CSE), Converged Security and Management Engine (CSME), and Global Security Controller (GSC) through the Management Engine Interface (MEI). It provides a unified, cross-platform API for interacting with MEI devices.

  • Direct Rendering Manager Library

    A user space library for accessing the Direct Rendering Manager (DRM). The library provides wrapper functions for the input/output control (ioctl) to avoid exposing the kernel interface directly.

  • Intel® RDMA Enablement Module

    A Linux kernel-mode driver that enables direct peer-to-peer communication between the memory of an Intel GPU supporting the Direct Memory Access Buffer (DMA-BUF) framework and an Host Channel Adapter (HCA) device with Mellanox’s PeerDirect interface.

  • GL Vendor-Neutral Dispatch Library

    A vendor-neutral dispatch layer that arbitrates OpenGL API calls across multiple vendors. It enables drivers from different vendors to coexist on the same filesystem and determines the appropriate vendor to handle each API call at runtime.

  • oneAPI Level Zero Ray Tracing Support

    A library that provides high-performance CPU-based construction algorithms for 3D acceleration structures, designed to work with the ray tracing hardware in Intel GPUs. Intel® oneAPI Level Zero uses this library to implement the Ray Tracing Acceleration Structures (RTAS) builder extension. The library is intended for use exclusively through Level Zero and is not meant to be accessed directly.

  • Video Decode and Presentation API for Unix

    A library for decoding, post-processing, compositing, and displaying both compressed and uncompressed video streams. It enables integration of video content with bitmap graphics for implementing on-screen displays (OSDs) and other application user interfaces.

  • Intel® Out-of-Band Management Services Module

    A module that provides an empty implementation of basic driver functionality for the function0 and function2 parts of Out-of-Band Management Services Module (OOBMSM) devices. This project is no longer under active development.