Introduction

Installing Intel® oneAPI Toolkits and Intel® Graphics Compute Runtime in HPC Cluster Environment

Download PDF

ID 766457

Date 11/07/2023

Version

Public

A newer version of this document is available. Customers should click here to go to the newest version.

Visible to Intel only — GUID: GUID-5A8E6735-9691-4FA1-BFC4-4F3D506A940D

Start with the Intel® oneAPI Base Toolkit, accompany it with the Intel® HPC Toolkit to enable Intel® MPI, Intel compilers with OpenMP* support, and Intel cluster diagnostic tools.

This guide uses a few terms that need to be explained before you proceed:

Intel® oneAPI toolkit is a collection of Intel® oneAPI packages that enable certain use cases:
- Intel® oneAPI Base Toolkit is the basis for Intel oneAPI. It provides the foundational toolset to develop performant, data-centric applications across Intel® CPUs, GPUs, and FPGAs.
- Intel® HPC Toolkit is an add-on to the Base toolkit. It provides the tools to build, analyze, and scale applications across shared- and distributed-memory computing systems.
- Other domain-specific toolkits are also available from the Intel oneAPI Toolkits page
Linux* kernel mode driver

Like any other hardware component, Intel GPUs require support by the Linux kernel. This support is provided by the Linux kernel mode driver (KMD).
Intel® Graphics Compute Runtime for oneAPI Level Zero and OpenCL™ Driver is an open source project providing compute API support (Level Zero, OpenCL) for Intel graphics hardware architectures (https://github.com/intel/compute-runtime).

Even though it is a collection of tools and shared libraries, it is also called user mode driver (UMD). OpenCL* uses the term driver and since there is a kernel mode driver, the term user mode driver is used.