PSTSWM Origin2000 Algorithm Comparison

Performance Studies using

PSTSWM

SGI/Cray Research Origin2000 Algorithm Comparison

(distributed LT experiment II-C )

Date/Person: April 22, 1999/ P. Worley
Platform: SGI Origin2000 at Los Alamos National Laboratory (n12):
   128 250-MHz MIPS R10000 processors
Environment: IRIX 6.5
mpt_1.3.0.0
MIPSpro Compilers: Version 7.2.1
setenv MPI_DSM_MUSTRUN 1
Code Version: 6.4.3
Compilation Options: f77 -64 -i4 -O3
Math Library: SCSL
Communication Library: MPI
SHMEM
Number of Timesteps: 12
Results:

Distributed LT (2) (mpi) / Column Major Processor Ordering
Algorithm Comparison
  T42L2     T42L2     T42L16    T85L4     T85L4     T85L32 
  P=2x32     P=4x16     P=8x8     P=2x32     P=4x16     P=8x8  
  optimal algorithm   halfsum  halfsum  halfsum  halfsum  halfsum  halfsum 
  (allreduce-min)/min     0.590    0.153    0.113    0.802    0.303    0.147 
  (generic-min)/min     1.552    0.655    0.038    0.071    0.071    0.012 

Distributed LT (2) (shmem) / Column Major Processor Ordering
Algorithm Comparison
  T42L2     T42L2     T42L16    T85L4     T85L4     T85L32 
  P=2x32     P=4x16     P=8x8     P=2x32     P=4x16     P=8x8  
  optimal algorithm   halfsum  halfsum  halfsum  halfsum  halfsum  halfsum 
  (generic-min)/min     0.402    0.219    0.064    0.240    0.149    0.067 

Distributed LT (2) (combined) / Column Major Processor Ordering
Communication Library Comparisons
  T42L2     T42L2     T42L16    T85L4     T85L4     T85L32 
  P=2x32     P=4x16     P=8x8     P=2x32     P=4x16     P=8x8  
  optimal library   shmem  shmem  shmem  shmem  shmem  shmem 
  (allreduce-min)/min     0.943    0.797    0.245    1.030    0.339    0.227 
  (mpi-min)/min     0.222    0.559    0.119    0.126    0.028    0.069 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:08:00 EDT.