PSTSWM SP3-200 Algorithm Comparison

Performance Studies using

PSTSWM

IBM SP3-200 Algorithm Comparison

(distributed LT experiment I-C )

Date/Person: June 6, 1999 / P. Worley
Platform: IBM SP3 at Oak Ridge National Laboratory (morgan.ccs.ornl.gov):
     62 2-way Winterhawk I SMP nodes (200 MHz POWER3 with 4MB L2 cache, equivalent to RS/6000 Model 260)
Environment: AIX 4.3.2;   PSSP 3.1
Code Version: 6.5
Compilation Options: mpxlf -O3 -qarch=auto -qtune=auto -qcache=auto
Link Options: -bmaxdata:0x70000000
Math Library: ESSL
Communication Library: MPI over the switch using user space for internode communication
MPI using IP for intranode communication
Number of Timesteps: 12
Results:

Distributed LT (1) (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  ringpipe  ringpipe  ringpipe 
  (allreduce-min)/min     0.988    0.375    0.059    0.434    0.184    0.093 
  (generic-min)/min     1.335    0.495    0.061    0.557    0.190    0.090 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:15:04 EDT.