[cig-commits] commit: Notes for parallel performance writing
Mercurial
hg at geodynamics.org
Tue May 15 15:57:08 PDT 2012
changeset: 125:3231eab98c1b
parent: 123:9a339f379222
user: Matthew G Knepley <knepley at gmail.com>
date: Tue May 15 18:56:14 2012 -0400
files: faultRup.tex
description:
Notes for parallel performance writing
diff -r 9a339f379222 -r 3231eab98c1b faultRup.tex
--- a/faultRup.tex Thu May 10 17:17:47 2012 -0400
+++ b/faultRup.tex Tue May 15 18:56:14 2012 -0400
@@ -1213,6 +1213,15 @@ number of iterations required for conver
number of iterations required for convergence.
\matt{Add additional comments, explanation of performance}
+Sources of load imbalance:
+ unstructured partition (1.3)
+ Dirichlet conditions (2.2)
+ fault unknowns (3)
+Reasons to eliminate unknowns
+
+
+Competition between integration and solve for balance
+
\subsection{Parallel Scaling Performance}
@@ -1257,6 +1266,10 @@ the number of iterates increases strongl
the number of iterates increases strongly with process count. The introduction of
Schur complement methods and AMG slows the growth considerably, but future work will
pursue the ultimate goal of iteration counts independent of process number.
+
+Separately plot VecMDot and MatMult
+This shows that increasing iterates have a penalty from sync VecMDot, and MatMult scales
+better. Could tradeoff with iterative method like BCGS.
% ------------------------------------------------------------------
\section{Code Verification Benchmarks}
More information about the CIG-COMMITS
mailing list