numpy performance reached on reasonable matrix size
This commit is contained in:
@@ -0,0 +1,3 @@
|
||||
gcc bench_dot.c ../src/lina.c -O3 -march=native -ffast-math -funroll-loops -o bench_dot
|
||||
./bench_dot
|
||||
python3 py_dot.py
|
||||
Reference in New Issue
Block a user