[Trilinos-Users] [EXTERNAL] Performance Study

[Rajamanickam, Sivasankaran]

Frey,

   The performance of Amesos2 will depend heavily on what is the underlying direct solver. The corresponding direct solver papers are your best reference. There is a small overhead for Amesos2, but it shouldn't be noticeable for large enough problems.


Performance of Belos has been documented in several application papers that use Trilinos. The solvers are typically what each application cared about. For example, see


http://www.sandia.gov/~srajama/publications/trilinos_sierra_performance.pdf

http://epubs.siam.org/doi/abs/10.1137/15M1040839


-Siva


________________________________
From: Trilinos-Users <trilinos-users-bounces at trilinos.org> on behalf of Frey Matthias (PSI) <matthias.frey at psi.ch>
Sent: Tuesday, February 20, 2018 4:41 AM
To: trilinos-users at trilinos.org
Subject: [EXTERNAL] [Trilinos-Users] Performance Study

Dear Trilinos-Users

I'm wondering about the performance (strong scaling and weak scaling) of several Trilinos packages, i.e. Belos and Amesos2. Unfortunately, I couldn't find any on the web.

Does anyone of you know where I can find some or has anyone of you already done a scaling study of solvers in Amesos2 or Belos?

Thanks for your help.

Best,
Matthias
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://trilinos.org/pipermail/trilinos-users/attachments/20180220/fea3887d/attachment.html>