Skip to content
This repository has been archived by the owner on Nov 8, 2022. It is now read-only.

Setting ParallelAccelerator to use MKL instead of OpenBLAS #151

Open
ig0xb opened this issue Jul 13, 2017 · 1 comment
Open

Setting ParallelAccelerator to use MKL instead of OpenBLAS #151

ig0xb opened this issue Jul 13, 2017 · 1 comment

Comments

@ig0xb
Copy link

ig0xb commented Jul 13, 2017

I was able to setup ParallelAccelerator to use OpenBLAS using this Link which i found in the previous post in the Issues section.
I tried to set up MKL_LIB=1 & OPENBLAS_LIB = 0 . But no luck i guess its more than that
I have also made the following changes to bashprofile

export CPLUS_INCLUDE_PATH=$CPLUS_INCLUDE_PATH:/opt/intel/mkl/include/
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/opt/intel/mkl/lib/

Kindly let me know what changes do i need to make to setup in order for ParallelAccelerator to use MKL.
Thank You
Regards
Saran

@DrTodd13
Copy link
Member

If you want to use MKL, I'd recommend using the following link to get Julia itself setup for MKL. Then, I'd re-build the ParallelAccelerator package with Pkg.build("ParallelAccelerator"). Look at generated/config.jl to make sure mkl_lib is set. If it isn't, let us know. You're approach of setting those directly in the config.jl would probably work but they are paths, not 1 or 0. So, I'd still suggest re-running the build script.

How to setup Julia for MKL.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants