[gmx-users] Gromacs on Gigabit Ethernet Dual Opteron Cluster
Ghermes at belozersky.msu.ru
Mon Oct 6 11:15:01 CEST 2003
>> The system (dummified protein in water)is big enough:~ 107.000 atoms,
>> -shuffle -sort were used.
BdG> was this using PME or cut-off? I guess it was cut-off, otherwise I'd be
BdG> surprised about the scaling of 80% for three nodes...
BdG> If you used cut-off, would you mind repeating the benchmark with PME?
Right, it was twin-range cut-off. Benchmarking PME is not
straightforward right now, since i had some problems with FFTW and
compiled Gromacs without this package. But we'll tune up fftw as soon
as other questions are clarified.
More information about the gromacs.org_gmx-users