Hello all,
I was repeating some of the benchmarking tests from
http://ambermd.org/gpus/benchmarks.htm on SDSC Comet,
and noticed that in “JAC_production_NVE_4fs” test with increased simulation time (100 ns),
both potential and kinetic energy linearly decreased by ~2000 kcal/mol each.
Temperature linearly decreased from 400K to 370K.
Is it expected behavior, and how could it be explained?
Some technical details:
| GPU (CUDA) Version of PMEMD in use: NVIDIA GPU IN USE.
| Version 16.0.0
| 02/25/2016
| Precision model in use:
| [SPFP] - Single Precision Forces, 64-bit Fixed Point
| Accumulation. (Default)
| CUDA_VISIBLE_DEVICES: 1
| CUDA Capable Devices Detected: 1
| CUDA Device ID in use: 0
| CUDA Device Name: Tesla K80
| CUDA Device Global Mem Size: 11439 MB
| CUDA Device Num Multiprocessors: 13
| CUDA Device Core Freq: 0.82 GHz
inpcrd, prmtop, mdin.GPU were from the original benchmarking tests.
The only modification was that ‘nstlim' was increased from 250,000 (i.e. 1 ns) to 25,000,000 (100 ns).
Thank you,
Marat
--
Marat R. Talipov
New Mexico State University
Department of Chemistry & Biochemistry
Las Cruces, NM 88003
Phone: 575.646.5210
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Jun 02 2017 - 08:00:03 PDT
