Specify number of threads for CPU core?

[jchodera]

Is there a way to specify number of threads for the CPU core? Does it automatically try to use all of them?

For one of our projects (91f51e8a-7b40-4adc-8ca9-38786a9fe654), the CPU ocore on 12 cores seems to be abysmally slow:

  date       time       tpf   ns/day  frames      steps
Jul/07 11:04:56PM  45:08:20     0.13       0         20 entering main md loop...
resuming from step 0

[kyleabeauchamp]

Not yet recall that feature was not included in 6.0.1
On Jul 7, 2014 11:21 PM, "John Chodera" [email protected] wrote:

Is there a way to specify number of threads for the CPU core? Does it
automatically try to use all of them?

For one of our projects (91f51e8a-7b40-4adc-8ca9-38786a9fe654), the CPU
ocore on 12 cores seems to be abysmally slow:

date time tpf ns/day frames steps
Jul/07 11:04:56PM 45:08:20 0.13 0 20 entering main md loop...
resuming from step 0

—
Reply to this email directly or view it on GitHub
#40.

[jchodera]

I think this means it tries to use all available threads instead, but 0.22 ns/day is godawful slow...

[proteneer]

It should try to use all cores. OpenMM is the only MD engine to exhibit
reverse scaling, ie. it gets slower the more threads you use :D

On Mon, Jul 7, 2014 at 8:25 PM, John Chodera [email protected]
wrote:

I think this means it tries to use all available threads instead, but 0.22
ns/day is godawful slow...

—
Reply to this email directly or view it on GitHub
#40 (comment).

Yutong Zhao
Stanford University

www.proteneer.com | simbios.stanford.edu