[Trilinos-Users] Determinism in iterative solvers (Tpetra, Belos)

[Christopher Thiele]

Hi all,

I am currently comparing a custom CG solver implementation against the CG
solver from the Belos package (Belos+Tpetra with OpenMP) to verify
correctness and to get an idea of its computational performance. When I
look at the residual norms in my solver, they are basically the same as
those I get from Belos, but with tiny deviations. These deviations are not
unexpected, as the CG method depends on dot products quite a lot and the
involved reductions are not fully deterministic, as the order in which
threads and processes sum up their intermediate results may vary.

However, I noticed that with the Belos solver the residuals were identical
every time, even up to the 25th digit or so. My question is how much of an
impact this deterministic behavior will have on the performance. It
probably requires additional synchronization, and I want to do a fair
comparison. Also, is there a way to disable deterministic reductions in
Trilinos/Tpetra/Kokkos altogether?

Thank you,
Christopher
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://trilinos.org/pipermail/trilinos-users/attachments/20160715/36329e1a/attachment.html>