PSTSWM Origin2000 Algorithm Comparison

Performance Studies using

PSTSWM

SGI/Cray Research Origin2000 Algorithm Comparison

(transpose LT experiment I-C )

Date/Person: May 11, 1998 / P. Worley
Platform: SGI Origin2000 at Los Alamos National Laboratory (nirvana11):
   128 195-MHz MIPS R10000 processors
Environment: IRIX 6.5
mpt_1.2.0.1
MIPSpro Compilers: Version MIPS pro_7.2
Code Version: 6.3
Compilation Options: f77 -64 -i4 -O3
Math Library: SCSL
Communication Library: MPI
SHMEM
(assuming distributed FFT)
Number of Timesteps: 12
Results:

Transpose LT (1) (mpi) / Column Major Processor Ordering
Algorithm Comparison
  T21L16    T42L16    T42L32    T42L16    T85L32 
  P=4x16     P=8x8     P=2x32     P=4x16     P=8x8  
  optimal algorithm   logtrans  srtrans  swtrans  srtrans  srtrans 
  (alltoallv-min)/min     0.104    0.052    1.503    0.357    0.158 
  (generic-min)/min     0.151    0.021    0.015    0.015    0.049 

Transpose LT (1) (shmem) / Column Major Processor Ordering
Algorithm Comparison
  T21L16    T42L16    T42L32    T42L16    T85L32 
  P=4x16     P=8x8     P=2x32     P=4x16     P=8x8  
  optimal algorithm   swtrans  swtrans  swtrans  swtrans  swtrans 
  (generic-min)/min     0.099    0.035    0.075    0.044    0.039 

Transpose LT (1) (combined) / Column Major Processor Ordering
Communication Library Comparisons
  T21L16    T42L16    T42L32    T42L16    T85L32 
  P=4x16     P=8x8     P=2x32     P=4x16     P=8x8  
  optimal library   shmem  mpi  shmem  mpi  mpi 
  (alltoallv-min)/min     0.566    0.052    1.511    0.357    0.158 
  (mpi-min)/min     0.419    0.000    0.003    0.000    0.000 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:07:28 EDT.