←back to thread

1080 points antipaul | 3 comments | | HN request time: 0.973s | source
1. madenine ◴[] No.25072732[source]
Is there a common linear algebra benchmark?

Would love to see a non-Intel chip + OpenBLAS beating out Intel + MKL

replies(1): >>25072769 #
2. kccqzy ◴[] No.25072769[source]
Why would you use OpenBLAS rather than Apple's Accelerate.framework? Since you're already using the Intel MKL, the only fair comparison is Accelerate.framework.
replies(1): >>25132670 #
3. moelf ◴[] No.25132670[source]
https://github.com/numpy/numpy/blob/master/doc/release/upcom...