[petsc-users] Computation to communication ratio for PETSc's routines

Alexander Grayver agrayver at gfz-potsdam.de
Wed Jun 13 05:45:10 CDT 2012


Hello,

As stated, I would like to estimate computation to communication ratio 
for PETSc's linear algebra routines (e.g. MatMult{Transpose}, MatPtAP, 
MatMatMult etc.)
As far as I understand the ratio depends on particular implementation 
and number of processes one runs application on.
So I guess these formulas should be known?
If not then I see one way to estimate it. Write app with those 
operations, parse -log_summary and then divide flops/messages, however 
the question is what information from -log_summary output should be used 
for that?

Thanks.

-- 
Regards,
Alexander



More information about the petsc-users mailing list