Hello, I've done some tests using the mom4 code. I have performed calculations on a rectangular basin with a 16x16x16 grid. We had done these type of calculations earlier using mom3. Now it seems that mom3 is about 4 times faster than mom4. Has anyone noticed this as well? I've done some experiments using perfex on an Origin 3000 system. mom4 seems to do twice as many floating point operations and suffers a lot from cache misses.
Has anyone suggestions what the cause of this difference may be? Some details of the calculation: - 16x16 4 degrees grid - Constant salt value - Temperature forcing applied. - No passive tracer - No fresh water flux - Free surface in both models (non linear explicit in MOM3). - Time steps the same for both with dtts = 43200 dteta = dtuv =200 and dtfs = 100 Regards, Fokke Dijkstra -- ----------------------------------------------------------------------- Fokke Dijkstra E-mail: [EMAIL PROTECTED] High Performance Computing Group Princetonplein 5 3584 CC Utrecht Phone: +31-30-2537716 ------------------------------------------------------------------------ ------------------------------------------------------- This SF.NET email is sponsored by: SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See! http://www.vasoftware.com _______________________________________________ FMS-mom4 mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/fms-mom4
