PSTSWM T3E-900 Algorithm Comparison

Performance Studies using

PSTSWM

SGI/Cray Research T3E-900 Algorithm Comparison

(distributed FFT experiment B)

Date/Person: May 11, 1998 / P. Worley
Platform: T3E-900 at National Energy Research Scientific Computing Center (mcurie.nersc.gov):
   532 450-MHz DEC Alpha EV5 RISC processors
Environment: UNICOS/mk 2.0.2.28
mpt 1.2.0.2
Cray CF90 Version 3.0.2.1
Code Version: 6.3
Compilation Options: -dp -Oscalar3 -Wl,-Dstreams=on
Math Library: libsci
Communication Library: MPI
SHMEM
Number of Timesteps: 12
Results:

Distributed FFT (mpi) / Row Major Processor Ordering
Algorithm Comparison
  T42L2     T42L2     T42L16    T85L4     T85L4     T85L32 
  P=32x2     P=16x4     P=8x8     P=32x2     P=16x4     P=8x8  
  optimal algorithm   dfft  generic  dfft  generic  dfft  dfft 
  (generic-min)/min     0.006    0.000    0.005    0.000    0.037    0.009 

Distributed FFT (shmem) / Row Major Processor Ordering
Algorithm Comparison
  T42L2     T42L2     T42L16    T85L4     T85L4     T85L32 
  P=32x2     P=16x4     P=8x8     P=32x2     P=16x4     P=8x8  
  optimal algorithm   dfft  dfft  dfft  dfft  dfft  dfft 
  (generic-min)/min     0.098    0.167    0.077    0.062    0.078    0.069 

Distributed FFT (combined) / Row Major Processor Ordering
Communication Library Comparisons
  T42L2     T42L2     T42L16    T85L4     T85L4     T85L32 
  P=32x2     P=16x4     P=8x8     P=32x2     P=16x4     P=8x8  
  optimal library   shmem  shmem  shmem  shmem  shmem  shmem 
  (mpi-min)/min     0.442    0.643    0.224    0.151    0.183    0.109 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:24:35 EDT.