This isn't exactly the benchmark info that you were looking for, but it
might be relevant:
I compared sander running with 4 processes and HT turned on to sander
running with 2 processes and HT turned off on a dual Xeon 2 GHz. Across
the 3 or 4 benchmarks I ran, the 4 process HT method was about 10-20%
slower.
This was with ifc/icc compiled sander from AMBER 8 circa July and LAM MPI.
John
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Wed Jan 14 2004 - 15:53:11 PST
