On 26. jan. 2015 12:59, Jørgen Kvalsvik wrote:
On 26. jan. 2015 12:47, [email protected] wrote:
Hello Jørgen,
great. If you have access to a super computer it would be very
interesting to know,
how the recently implemented new versions of GMRes and CG (that
implement overlapping of
gobal reductions with matrix-vector multiplications) perform in
comparison to the standard versions in terms of scaling. These
algorithms are in PETSc 3.4 and later if I remember correctly.
That's just a suggestion though.
Best,
Robert
I do have access to a recently-aquired numascale-machine, which I will
run something on if I find the time.
_______________________________________________
Opm mailing list
[email protected]
http://www.opm-project.org/mailman/listinfo/opm
