Half GETRF time is higher than Double GETRF

11 views
Skip to first unread message

Nima Sahraneshin

unread,
Sep 17, 2020, 3:56:25 PM9/17/20
to MAGMA User
Hi, 

I am working with MAGMA on Volta GPU and its strange for me to see the time of running hgetrf is higher than dgetrf. I have tested the same code on P100 and it was OK but here I am seeing something else. 

Do you have any recommendations? What is the most accurate method of timing? 

Best regards, Nima

Mark Gates

unread,
Sep 17, 2020, 4:23:20 PM9/17/20
to Nima Sahraneshin, MAGMA User
Can you give a specific example, so we can understand and replicate your results? Knowing your machine configuration (e.g., CPU, # cores/threads, GPU, BLAS/LAPACK library, CUDA version) is helpful. Especially if you can provide an example using MAGMA's testers, that is helpful. Please include complete input & output, as that provides valuable information, including some of the above.

Mark

--
You received this message because you are subscribed to the Google Groups "MAGMA User" group.
To unsubscribe from this group and stop receiving emails from it, send an email to magma-user+...@icl.utk.edu.
To view this discussion on the web visit https://groups.google.com/a/icl.utk.edu/d/msgid/magma-user/CAD%3DWvJb4yCRWoE7xE%2Bxvd8nwiXpVJzkiKrYSOUU0FhJ%3DPOBM1g%40mail.gmail.com.


--
Innovative Computing Laboratory
University of Tennessee, Knoxville
Reply all
Reply to author
Forward
0 new messages