Arkansas High Performace Computing Center [hpcwiki]

User Tools

Site Tools

Trace:
namd

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
namd [2020/11/20 20:30] – rootnamd [2025/10/15 19:51] (current) – external edit 127.0.0.1
Line 4: Line 4:
  
 This is for multiple-node runs with ''charmrun'' as the distributed component and ''namd2'' on each compute node. We have found most runs are faster with the ''+setcpuaffinity +isomalloc_sync'' options. charmrun ''++ppn'' should match PBS ''ppn=''. This is for multiple-node runs with ''charmrun'' as the distributed component and ''namd2'' on each compute node. We have found most runs are faster with the ''+setcpuaffinity +isomalloc_sync'' options. charmrun ''++ppn'' should match PBS ''ppn=''.
 +
 <code> <code>
 module load namd/2.12  [or 2.11] module load namd/2.12  [or 2.11]
Line 46: Line 47:
 16 Intel Razor    16      2.12          1   242 16 Intel Razor    16      2.12          1   242
 32 AMD Trestles   32      2.12          1   315 32 AMD Trestles   32      2.12          1   315
-32 Intel Pinnacle 32      2.12          1   127 +32 Intel G6130    32      2.12          1   127 
-32 Intel Pinnacle 32      2.13          1   127 +32 Intel G6130    32      2.13          1   127 
-32 Intel Pinnacle 32      2.15a1-AVX512 1    76 +48 AMD Epyc 7402  48      2.13          1    89 
-32 Intel Pinnacle 32 V100 3.0a7-cuda    1    39 +32 Intel G6130    32      2.15a1-AVX512 1    76 
-48 AMD Epyc 2     48      2.15a1-AVX2 needs recompilation+32 Intel G6130    32 V100 3.0a7-cuda    1    39 
 +48 AMD Epyc 7402  48      2.15a1-AVX2 needs recompilation
 </code> </code>
  
  
  
namd.1605904223.txt.gz · Last modified: (external edit)