HeadCrab

GPU621/DPS921 | Participants | Groups and Projects | Resources | Glossary

Intel Math Kernel Library (MKL)

Team Member

1. Rene Anderson

Intro

For this assignment I will be focusing on the Linear Algebra routines. I will use workshop six to demonstrate how BLAS can be used to significantly speed up the calculations.

Compare speeds to

1. Serial
2. Cilk
3. Cilk with array notation and reduction
4. Cilk with SIMD
5. MKL CBLAS level 3
6. CUDA CUBLAS level 3

Intel Math Kernel Library

MKL provides highly vectorized and threaded Linear Algebra, Fast Fourier Transforms, Vector Math and Statistics functions. Intel MKL uses industry standard APIs. This means that developers would have to make minor changes to their programs when switching to MKL.

Intel MKL gives the developer control over the necessary trade-offs

1. Result consistency vs performance
2. Accuracy vs performance

Intel MKL is also compatible with your choice of compilers, languages, operating systems, linking and threading models. One library solution across multiple environments means only one library to learn and manage.

Linear Algebra

Intel MKL provides highly optimized BLAS routines

1. BLAS Level 1 vector-vector
2. BLAS Level 2 matrix-vector
3. BLAS Level 3 matrix-matrix

Source code

How to enable Intel MKL

Command line

1. -mkl
2. -mkl=parallel to link with standard threaded Intel MKL.
3. -mkl=sequential to link with sequential version of Intel MKL.
4. -mkl=cluster to link with Intel MKL cluster components (sequential) that use Intel MPI.

Microsoft Visual Studio

project properties->Intel Performance Libraries->Intel Math Kernel Library

Useful Link

1. https://software.intel.com/en-us/intel-mkl/details
2. https://software.intel.com/en-us/node/468380
3. https://software.intel.com/sites/default/files/managed/4a/d6/mkl_11.2.1_lnx_userguide.pdf