[CSC 435] OpenBLAS timing.

[Andrew J. Pounds]

I think some of you are confused about the timing project.  Everything
is done with 9000x9000 matrices.

You need to use *ONE* system that is in the ithaca cluster for...

1. GPU timing with the CudaBLAS  (run 10 times in CUDA)

2. Use your own OpenMP Matrix Multiply code run on 1-8 threads (10 times
each)

Then on HAMMER

3. Use your own OpenMP Matrix Multiply code run on 1-40 cores (10 times
each)

4. Using the OpenBLAS Matrix Multiply code run on 1-40 cores (10 times each)


You are welcome to use the batch job code I gave you.  I class tomorrow
maybe we will have time to complete some of these items.



-- 
Andrew J. Pounds, Ph.D.  (pounds_aj at mercer.edu)
Professor of Chemistry and Computer Science
Mercer University,  Macon, GA 31207   (478) 301-5627
http://faculty.mercer.edu/pounds_aj

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://theochem.mercer.edu/pipermail/csc435/attachments/20180423/d04440d0/attachment.html>