4 ms·
Good writeup and commendable of you to make your benchmark so easily repeatable. On my 16-core Xeon(R) W-2245 CPU @ 3.90GHz matmul.c takes about 1.41 seconds to
by bjourne 2y ago
Good writeup and commendable of you to make your benchmark so easily repeatable. On my 16-core Xeon(R) W-2245 CPU @ 3.90GHz matmul.c takes about 1.41 seconds to multiply 8192x8192 matrices when compiled with gcc -O3 and 1.47 seconds when compiled with clang -O2, while NumPy does it in 1.07 seconds. I believe an avx512 kernel would be significantly faster. Another reason for the lackluster performance may be omp. IME, you can reduce overhead by managing the thread pool explicitly with pthreads (and use sysconf(_SC_NPROCESSORS_ONLN) instead of hard-coding).