You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by GitBox <gi...@apache.org> on 2019/11/21 00:37:51 UTC

[GitHub] [incubator-tvm] cylinbao edited a comment on issue #4369: [Runtime] Add cusparse for sparse dense

cylinbao edited a comment on issue #4369: [Runtime] Add cusparse for sparse dense
URL: https://github.com/apache/incubator-tvm/pull/4369#issuecomment-556636358
 
 
   Thank @Huyuwei for sharing the interesting profiling results!
   Just want to know more about the details.
   The profiling include the time of the extra transpose on B to make the results the same, right?
   and I think it will be good to have another pr to apply the transpose(B) trick and use csrmm2().
   
   @yuluny2 
   The thing is the replacement function cusparseSpMM() only comes with the very new CUDA releases. (10.1.2 if I remember right). I think it's safe to use csrmm() or csrmm2() for now, and change to SpMM() later.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services