Developer Guide

Developer Guide for Intel® oneAPI Math Kernel Library macOS*

ID 766688
Date 12/16/2022
Public

A newer version of this document is available.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

Document Table of Contents

Version Information Line

In the Intel® oneAPI Math Kernel Library Verbose mode, the first call to a verbose-enabled function prints a version information line. The line begins with theMKL_VERBOSE character string and uses spaces as delimiters. The format of the rest of the line may change in a future release.

The following table lists information contained in a version information line and provides available links for more information:

Information

Description

Related Links

Intel® oneAPI Math Kernel Library version.

This information is separated by a comma from the rest of the line.

 

Operating system.

Possible values:

  • Lnx for Linux* OS
  • Win for Windows* OS
  • OSX for macOS*
 

The host CPU frequency.

   

Intel® oneAPI Math Kernel Library interface layer used by the application.

Possible values:

  • lp64 or ilp64 on systems based on the Intel® 64 architecture.

Using the ILP64 Interface vs. LP64 Interface

Intel® oneAPI Math Kernel Library threading layer used by the application.

Possible values:

intel_thread, tbb_thread, or sequential.

Linking with Threading Libraries

The following is an example of a version information line:

MKL_VERBOSE Intel(R) MKL 11.2 Beta build 20131126 for Intel(R) 64 architecture Intel(R) Advanced Vector Extensions (Intel(R) AVX) Enabled Processor, OSX 3.10GHz lp64 intel_thread

GPU Information Line

In Intel® oneAPI Math Kernel Library Verbose mode for GPU applications, the first call to a verbose-enabled function prints out the GPU information line or lines for all detected GPU devices, each in a separate line. The line begins with the MKL_VERBOSE Detected character string and uses spaces as delimiters. The format of the rest of the line may change in a future release.

Only Intel® GPU is supported.

The following table lists information contained in a GPU information line.

Information Description
GPU index The index of the GPU device will be printed after the character string "GPU" (e.g. GPU0, GPU1, GPU2, etc). This GPU index will be used as a nickname of the device in call description lines to refer to the device.
Intel® GPU architecture

The value can be one of the following:

  • Intel(R) Gen9
  • Intel(R) Xe_LP
  • Intel(R) Xe_HP
  • Intel(R) Xe_HPG
  • Intel(R) Xe_HPC
  • Unknown GPU
Runtime backend The value printed is prefixed with Backend:
Vector Engine number The value printed is prefixed with VE:
Stack number The value printed is prefixed with Stack:
Maximum workgroup size The value printed is prefixed with maxWGsize:

The following is an example of a GPU information line:

MKL_VERBOSE Detected GPU0 Intel(R)_Gen9 Backend:OpenCL VE:72 Stack:1 maxWGsize:256