GENESIS

Benchmark

Benchmark of GENESIS 1.3 and 1.4 on PC clusters

Force field: CHARMM
Integrator : VVER  (PME every 2 steps)
Ensemble : NVE
Time step : 2 fs

CPU: Intel Xeon Gold 6142 (2.6 GHz)
GPU: Geforce GTX 1080 Ti
Precision : Mixed precision

DHFR (23,558 atoms) switch, cutoff, pairlist = 10, 12, 13.5 Å
STMV (1,066,628 atoms) switch, cutoff, pairlist= 10, 12, 13.5 Å

Benchmark of GENESIS 1.0 and 1.1

Integrator : VVER or RESPA (PME every 2 steps)
Ensemble : NVE
Time step : 2 fs

• Intel and GPU benchmarks (CPU, CPU+GPU)
  CPU: Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz (2 CPU x 8 cores/node)
  GPU: NVIDIA Tesla K20x (2 GPUs/node)
  Compiler: openmpi 1.10.1 + intel (fortran/C) 15.0.5.223 + cuda 6.5
  Precision : double precision (dp) or mixed precision (sp)
• K computer benchmarks
  CPU: SPARC64(TM) VIIIFx (8 cores/node)
  Compiler: Fujitsu compiler
  Precision: double precision

MSPA (216,726 atoms)  on Intel CHARMM c36FF PME grid = 144,144,144 switch, cutoff, pairlist= 10, 12, 13.5
STMV (1,066,628 atoms)  on Intel CHARMM FF PME grid = 256, 256, 256 switch, cutoff, pairlist = 9, 10, 11.5
27 x STMV (28,798,956 atoms)  on K CHARMM FF PME grid = 512, 512, 512 switch, cutoff, pairlist = 9, 10, 11.5 Å