The Intel MKL library

⚠ INFORMATION
This page was translated by an AI (LLM) with a cursory human check and is awaiting full review.

Description

Intel MKL (Math Kernel Library) is a library that offers highly optimised sequential and parallel mathematical routines. The basic mathematical functions include BLAS, LAPACK, ScaLAPACK, BLACS, sparse solvers and FFT.

Installed versions

The versions installed on Jean Zay can be found using the following command:

module avail intel-mkl

Environment

For the module command, the versions of Intel MKL do not depend on any environment.

However, we recommend using each version of the Intel MKL library with the corresponding version of the Intel compiler (see the modules intel-compiler/...).

To obtain information on the library access paths, simply use the following command:

module show intel-mkl

Usage

The library is accessible via module.

To load the default version into the current environment:

module load intel-mkl

or to load a specific version:

module load intel-mkl/19.0.5

warning

For the module command, the versions of Intel MKL do not depend on any environment. Therefore, no compiler version will be automatically loaded when activating a version of Intel MKL. We recommend using each version of the Intel MKL library with the corresponding version of the Intel compiler (see module avail intel-compilers).

Link editing

Link editing for the use of MKL can vary depending on several parameters such as the operating system, the type of compiler and the architecture. It is recommended to use Intel® MKL Link Line Advisor to find the appropriate options.

note

The following examples use the default Intel MKL library and the default Intel compiler to ensure a consistent environment.

Here is an example for the BLAS library:

module load intel-mklmodule load intel-compilersifort appel_blas.f90 -L${MKLROOT}/lib/intel64 -lmkl_intel_lp64 \ 
 -lmkl_sequential -lmkl_core -lpthread -lm -ldl

Here is an example for the LAPACK library:

module load intel-mklmodule load intel-compilersifort appel_lapack.f90 -L${MKLROOT}/lib/intel64 -lmkl_intel_lp64 \ 
 -lmkl_sequential -lmkl_core -lpthread -lm -ldl

Here is an example for the ScaLAPACK library:

module load intel-mklmodule load intel-compilersifort appel_scalapack.f90 -L${MKLROOT}/lib/intel64 \ 
 -lmkl_scalapack_lp64 -lmkl_blacs_intelmpi_lp64 \ 
 -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread -lm

Documentation

📝 Intel MKL documentation