large matrix core dump

[cccpocker]

my device is a cpu 32 cores and memory 64g machine. The matrix shape is (2000000,512), and when i use these to get result, it makes core dump.

HREADS exceeded, adding auxiliary array for thread metadata.
OpenBLAS warning: precompiled NUM_THREADS exceeded, adding auxiliary array for thread metadata.
Segmentation fault (core dumped)

[rragundez]

Hi @cccpocker thanks for raising the issue. Seems like an OpenBLAS bug, you can read here: https://stackoverflow.com/questions/75619847/python-sklearn-openblas-error-for-kmeans and https://github.com/OpenMathLib/OpenBLAS/issues/3321

Can you post the full stack trace, in which line in the code does this happen?

[rragundez]

Hi @cccpocker did you manage to find the solution? I hope so!