following guidelines from https://github.com/flame/how-to-optimize-gemm
$ cd cpu
$ ./batch_run.sh
$ python ./plot_all.py
10~20 times faster than the orignial version(compiled by gcc with -O2
flag)
following guidelines from https://github.com/flame/how-to-optimize-gemm
$ cd cpu
$ ./batch_run.sh
$ python ./plot_all.py
10~20 times faster than the orignial version(compiled by gcc with -O2
flag)