ICLDisco / dplasma

DPLASMA is a highly optimized, accelerator-aware, implementation of a dense linear algebra package for distributed heterogeneous systems. It is designed to deliver sustained performance for distributed systems where each node featuring multiple sockets of multicore processors, and if available, accelerators, using the PaRSEC runtime as a backend.
Other
10 stars 8 forks source link

bugfix: divide by zero in zgemm gpu #77

Closed abouteiller closed 1 year ago

abouteiller commented 1 year ago

fix divide by zero in zgemm wrapper (as seen during ctest- zgemm-lowmem) fix Zgemm function name not being converted (harmless but alarming) reset device loads in zgemm testers between loads reduce memory usage during zgemm testers