High-performance computing
(HPC) is at the core of AI, machine learning, and deep learning
applications. The Intel¢ç oneAPI HPC Toolkit (HPC Kit) delivers what
developers need to build, analyze, optimize, and scale HPC applications
with the latest techniques in vectorization, multithreading, multi-node
parallelization, and memory optimization.
This toolkit is an add-on
to the Intel¢ç oneAPI Base Toolkit, which is required for full
functionality. It includes powerful data-centric libraries, advanced
analysis tools, and Intel¢ç Distribution for Python* for near-native code
performance of core Python numerical, scientific, and machine learning
packages.
oneAPI BASE Toolkit
oneAPI BASE & HPC Toolkit
oneAPI BASE & IOT Toolkit
Intel¢ç oneAPI Collective Communications
Library
✔
✔
✔
Intel¢ç oneAPI Data Analytics Library
✔
✔
✔
Intel¢ç oneAPI DPC++/C++ Compiler
✔
✔
✔
Intel¢ç oneAPI DPC++ Library
✔
✔
✔
Intel¢ç oneAPI Deep Neural Network Library
✔
✔
✔
Intel¢ç oneAPI Math Kernel Library
✔
✔
✔
Intel¢ç oneAPI Threading Building Blocks
✔
✔
✔
Intel¢ç oneAPI Video Processing Library
✔
✔
✔
Intel¢ç Advisor
✔
✔
✔
Intel¢ç Distribution for Python*
✔
✔
✔
Intel¢ç DPC++ Compatibility Tool
✔
✔
✔
Intel¢ç Integrated Performance Primitives
✔
✔
✔
Intel¢ç VTune¢â Profiler
✔
✔
✔
Intel¢ç FPGA Add-On for oneAPI Base
Toolkit (Separate Download Required)
✔
✔
✔
Intel¢ç Distribution for GDB*
✔
✔
✔
Intel¢ç oneAPI DPC++/C++ Compiler
✔
✔
Intel¢ç C++ Compiler Classic
¡¡
✔
✔
Intel¢çCluster Checker
✔
Intel¢ç Fortran Compiler
¡¡
✔
¡¡
Intel¢ç Fortran Compiler Classic
✔
Intel¢ç Inspector
¡¡
✔
✔
Intel¢ç MPI Library
✔
Intel¢ç Trace Analyzer and Collector
¡¡
✔
¡¡
OpenEmbedded meta-intel Layer for Yocto Project*
¡¡
✔
Intel¢ç Embree
¡¡
¡¡
¡¡
Intel¢ç Implicit SPMD Program Compiler
¡¡
Intel¢ç Open Image Denoise
¡¡
¡¡
¡¡
Intel¢ç Open Path Guiding Library (Intel¢ç Open
PGL)
¡¡
Intel¢ç Open Volume Kernel Library (Intel¢ç Open
VKL)
¡¡
¡¡
¡¡
Intel¢ç OpenSWR
¡¡
Intel¢ç OSPRay
¡¡
¡¡
¡¡
Intel¢ç OSPRay for Hydra*
¡¡
Intel¢ç OSPRay Studio
¡¡
¡¡
¡¡
Rendering Toolkit Utilities
¡¡
Common Hardware Requirements
CPUs:
Systems based on
Intel¢ç 64 architectures below are supported both as host and target
platforms.
~3 GB of disk space (minimum)
if only installing compiler and its libraries: Intel oneAPI DPC++/C++
Compiler, Intel¢ç DPC++ Compatibility Tool, Intel¢ç oneAPI DPC++ Library
and Intel¢ç Threading Building Block
Maximum of ~24 GB diskspace
if installing all components
NOTE: During the
installation process, the installer may need up to 6 GB of additional
temporary disk storage to manage the download and intermediate installation
files.
The operating systems
listed below are supported on Intel¢ç 64 Architecture. Individual tools may
support additional operating systems and architecture configurations. See
the individual tool release notes for full details.
For developing
applications for offloading to accelerators like GPU or FPGA, a specific
version of GPU driver is required for the supported operating system.
Please visit Installation
Guide for Intel¢ç oneAPI Toolkits "Install
Intel GPU Drivers" section for up to date information.
Intel¢ç oneAPI 2023.2
is validated on Windows and Linux for the following driver versions
Windows Intel¢ç
Graphics Driver
To install the driver follow the directions in the article appropriate
for your device
Intel¢ç Arc¢â Graphics,
11th-13th Gen Intel¢ç Core¢â processor directions.
Xe Dedicated, 6th-10th Gen
Intel¢ç Core¢â Processor Graphics, and related Intel Atom¢ç, Pentium¢ç,
and Celeron¢ç processors directions.
Driver version varies depending on the Intel Graphics in the system.
Intel¢ç Data Center GPU Flex
Series (ATS-M). Contact your OEM representative for access to
the Intel Registration Center.
Note: These OS
distributions are tested by Intel or known to work; other distributions may
or may not work and are not recommended. If you have questions, access the Intel Community
Forums when you need assistance. If you have Commercial
Support, create a support ticket.
For
Linux*
GNU* Bash is
required for local installation and for setting up the environment to
use the toolkit.
Intel¢ç oneAPI tools
support for integrated graphics processors is available on 11th Generation
and newer Intel¢ç Core¢â Processors. Active support for older integrated
graphics processors included with 6th through 10th Generation Intel Core
Processors and related Intel Atom¢ç, Pentium¢ç, and Celeron¢ç processors will
no longer be available starting with the Intel oneAPI 2023.1 release.
For Windows*
For CPU
Support
OS (for CPU)
Supported Components
Windows* Pro & Enterprise 10, 11
All except oneCCL
Windows Server 2019*, 2022*
All except oneCCL
For GPU
Accelerator Support for iGPU
OS (for GPU)
Supported Components
Windows* 10, 11
All except oneCCL & oneTBB
Windows Server 2019*, 2022*
All except oneCCL & oneTBB
Intel¢ç oneAPI tools
support for integrated graphics processors is available on 11th Generation
and newer Intel¢ç Core¢â Processors. Active support for older integrated
graphics processors included with 6th through 10th Generation Intel Core
Processors and related Intel Atom¢ç, Pentium¢ç, and Celeron¢ç processors will
no longer be available starting with the Intel¢ç oneAPI 2023.1 release.
Note:
- macOS* support for Mac* computers with Intel¢ç Processors.
- Intel oneAPI Base Toolkit for macOS on x86 is now deprecated and will be
discontinued in the 2024.0 release. Several Intel-led open source
developer tool projects will continue supporting macOS on Apple Silicon
including oneAPI Threading Building Blocks (oneTBB) and Intel¢ç Implicit
SPMD Program Compiler and we welcome the opportunity to work with
contributors to expand support to additional tools in the future.