PSTSWM SP3-375 Communication Protocol Performance

Performance Studies using

PSTSWM

IBM SP3-375 Winterhawk II Protocol Performance

(transpose LT experiment I-A / O(log P) transpose algorithm)

Date/Person: December 12, 1999 / P. Worley
Platform: IBM SP3 at Oak Ridge National Laboratory (bobcat.ccs.ornl.gov):
     62 4-way Winterhawk II SMP nodes (375 MHz POWER3 with 8MB L2 cache)
Environment: AIX 4.3.3;   PSSP 3.1.1
Code Version: 6.6.4
Compilation Options: mpxlf -O3 -qarch=auto -qtune=auto -qcache=auto
Link Options: -bmaxdata:0x70000000
Environment Variables: export MP_SHARED_MEMORY="yes"
Math Library: ESSL
Communication Library: MPI over the switch using user space for internode communication
MPI using shared memory for intranode communication
Parallel Algorithm: logtrans
(assuming distributed FFT)
Number of Timesteps: 12
Results:

1x16 Processors / Problem T10L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.0875e-01   0.11   0.09   0.20 
Three Fastest
Protocols
1st2nd3rd
  c4   c5   d4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   4   25 

1x8 Processors / Problem T21L8
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.1712e-01   0.08   0.08   0.16 
Three Fastest
Protocols
1st2nd3rd
  c4   c5   c2 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   5   25 

1x16 Processors / Problem T21L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  2.0530e-01   0.09   0.07   0.16 
Three Fastest
Protocols
1st2nd3rd
  c4   c5   c2 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   5   25 

1x32 Processors / Problem T21L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.4138e-01   0.08   0.07   0.15 
Three Fastest
Protocols
1st2nd3rd
  c4   c5   a4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   7   25 

1x8 Processors / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  8.1025e-01   0.04   0.04   0.09 
Three Fastest
Protocols
1st2nd3rd
  c4   c2   c5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  7   14   25 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:19:18 EDT.