gilgamesh: MD benchmarks

[Craig Maloney]

Hi all.

Here are some MD benchmarks for gilgamesh.  These are LAMMPS runs with  
LJ-interactions in a 3D cube with PBCs on all sides.   As a rule of  
thumb, it looks like it's efficient to run a multi-node job with as  
few as 4000 atoms per node, but no fewer.  The numbers will be a bit  
different in 2D.

Have fun with the new machine.

--Craig


-----------------------------------------------
Contact info:
http://www.ce.cmu.edu/People/faculty/maloney.html



-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://lists.andrew.cmu.edu/mailman/private/gilgamesh-users/attachments/20100805/5f2a8b3b/attachment.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ljMelt-bench.pdf
Type: application/pdf
Size: 5435 bytes
Desc: not available
Url : https://lists.andrew.cmu.edu/mailman/private/gilgamesh-users/attachments/20100805/5f2a8b3b/attachment.pdf 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://lists.andrew.cmu.edu/mailman/private/gilgamesh-users/attachments/20100805/5f2a8b3b/attachment-0001.html