memory use on mkl in clusters

Nelson__Donald — Fri, 03 May 2019 15:56:10 GMT

Hello,

There are some very impressive memory vs mpi process plots in the excellent mkl presentation:

https://cerfacs.fr/wp-content/uploads/2016/03/Kalinkin.pdf

but its a little confusing what the memory requirements are, is the original matrix needed on each node? Sounds like it is from :
https://software.intel.com/en-us/mkl-developer-reference-fortran-dss-distributed-symmetric-matrix-storage
"The algorithm ensures that the memory required to keep internal data on each MPI process is decreased when the number of MPI processes in a run increases. However, the solver requires that matrix A and some other internal arrays completely fit into the memory of each MPI process."

Any thoughts appreciated, thanks!

Don

Hi Donald,

Alexander_K_Intel2 — Thu, 09 May 2019 20:35:12 GMT

Hi Donald,

You are right, in presentation which you mentioned initial matrix stored on master process. However in last releases of MKL we implemented distributed reordering ( iparm[1] = 10 ) which doesn't combine matrix on one processor.

Thanks,

Alex

topic memory use on mkl in clusters in Intel® oneAPI Math Kernel Library

memory use on mkl in clusters

Hi Donald,