Github user VinceShieh commented on the issue:
https://github.com/apache/spark/pull/18936
thanks, Sean and Nick.
To @srowen , I think the difference is the finding from our previous
investigation that, thread setting in the native BLAS impacts the overall
performance of a method/algorithm.
To @MLnick, Agree. We know it demands a certain amount of benchmark work
for this PR, since the changes are in a low level of the stack, they will
impact several methods and there are also other native BLAS implementations,
not just MKL, So, we take SVM as an example to show what we might get from it.
Also, given the fact that mllib is only in maintenance mode, pls let us know if
such change is unworthy of the work required.
Thanks!
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
