search

explosion / cython-blis

💥 Fast matrix-multiplication as a self-contained Python library – no system dependencies!
Other
219 stars 37 forks source link

Update to the latest flame/blis? #29

Closed rhazegh closed 4 years ago

rhazegh commented 4 years ago

I am having performance issues with this library on my AMD Ryzen 3900X CPU and was looking in the code to see if I can find out why this is happening. See this issue for reference: https://github.com/conda-forge/cython-blis-feedstock/issues/18

I have installed this through conda forge:

# conda list | grep blis
cython-blis               0.4.1            py38h1e0a361_1    conda-forge

It looks like cython-blis is using a version of flame/blis from Mar 12, 2019 (https://github.com/honnibal/blis/tree/2c59c46b57cb3517d33b350cc5d18022e1fe93f4)

It appears that since then, some new optimizations are added for AMD in the flame/blis library and they have added support for zen2 (https://github.com/flame/blis/tree/master/config/zen2)

Is it possible to update cython-blis to take advantage of these new optimizations?

honnibal commented 4 years ago

Currently working on this, thanks for your patience.

rhazegh commented 4 years ago

@honnibal Thank you. Please let me know if there is anything I can help with.

honnibal commented 4 years ago

v0.7.0 now up. Working on the conda package.

rhazegh commented 4 years ago

@honnibal Thank you very much for making this happen. Is there any chance we could also get a 0.7 conda package? If there is anything you think I can help with please let me know.