[Bug 1177260] Switching python-numpy to plain BLAS causes significant performance regression
https://bugzilla.suse.com/show_bug.cgi?id=1177260 https://bugzilla.suse.com/show_bug.cgi?id=1177260#c41 --- Comment #41 from Stefan Br�ns <stefan.bruens@rwth-aachen.de> --- (In reply to Egbert Eich from comment #39)
Ok, then the reason why the CBLAS-enabled numpy version was lost was that Leap acquired the python packages from SLE - where we obviously weren't able to build against it. The first step towards this would be to update Lapack on SLE to a version that provides CBLAS.
Originally, numpy used the cblas from openblas, so this did not matter. Only when numpy changed from openblas to netlib-lapack (@3.5.0), the missing cblas became a problem. For numpy + cblas, there are two options: 1. Build numpy as part of Backports, which already *has* cblas. 2. Bring an updated lapack (>= 3.6.0) to SLE. But as far as I understand, this *also* goes into backports only, so we also need to build numpy in backports. -- You are receiving this mail because: You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply@suse.com