PSTSWM SP3-200 Communication Protocol Performance

Performance Studies using

PSTSWM

IBM SP3-200 Protocol Performance

(transpose LT experiment I-C / O(P) sendrecv transpose algorithm)

Date/Person: June 8, 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
Parallel Algorithm: srtrans
(assuming distributed FFT)
Number of Timesteps: 12
Results:

4x16 Processors (Column Major Ordering) / Problem T21L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.8339e-01   0.25   0.22   0.51 
Three Fastest
Protocols
1st2nd3rd
  e3   e4   e5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   3   17 

2x32 Processors (Column Major Ordering) / Problem T21L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  2.1708e-01   0.37   0.37   0.76 
Three Fastest
Protocols
1st2nd3rd
  e4   e3   e5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   3   7 

8x8 Processors (Column Major Ordering) / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  5.2174e-01   0.08   0.09   0.24 
Three Fastest
Protocols
1st2nd3rd
  c4   e5   c5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   10   31 

4x16 Processors (Column Major Ordering) / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.6617e-01   0.15   0.12   0.29 
Three Fastest
Protocols
1st2nd3rd
  e5   e3   e4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   4   28 

2x32 Processors (Column Major Ordering) / Problem T42L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  6.7577e-01   0.26   0.28   0.56 
Three Fastest
Protocols
1st2nd3rd
  e4   e5   c4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   2   14 

8x8 Processors (Column Major Ordering) / Problem T85L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.0128e+00   0.06   0.05   0.28 
Three Fastest
Protocols
1st2nd3rd
  a1   c4   a5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   16   30 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:16:52 EDT.