PSTSWM Parallel Performance

Performance Studies using

PSTSWM

Experiment A Distributed FFT Performance

These results represent the minimum execution time for each platform, using the optimal parallel implementation, math libraries where available, and the best available compilers. The legends list the platforms in alphabetical order. Results are presented for both MPI-only experiments and best over all communication libraries.

Distributed FFT (mpi)
Platform Comparison
  T42L1     T21L2     T42L2     T85L2     T85L1     T85L4  
  P=32     P=16     P=8     P=32     P=16     P=8  
  best platform   sp3-200-nn  sp3-200-nn  or2k-250  sp3-200-nn  sp3-200-nn  sp3-200-nn 
  best timing   .56003E-01  .39443E-01  0.22927E+00  .35521E+00  .26860E+00  .19613E+01 

Distributed FFT (all)
Platform Comparison
  T42L1     T21L2     T42L2     T85L2     T85L1     T85L4  
  P=32     P=16     P=8     P=32     P=16     P=8  
  best platform   t3e900  t3e900  or2k-250  t3e900  sp3-200-nn  sp3-200-nn 
  best library   shmem  shmem  shmem  shmem  mpi  mpi 
  best timing   0.43247E-01  0.35001E-01  0.20973E+00  0.34754E+00  .26860E+00  .19613E+01 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 09:58:50 EDT.
81909 accesses since 1/2/96.