PSTSWM SP3-200 Communication Protocol Performance

Performance Studies using

PSTSWM

IBM SP3-200 Protocol Performance

(transpose LT experiment I-C / O(P) swap 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: swtrans
(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
  1.8358e-01   0.23   0.21   0.45 
Three Fastest
Protocols
1st2nd3rd
  e5   e4   e3 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   3   18 

2x32 Processors (Column Major Ordering) / Problem T21L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  2.1220e-01   0.41   0.41   0.89 
Three Fastest
Protocols
1st2nd3rd
  e4   e5   e3 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   3   8 

8x8 Processors (Column Major Ordering) / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  5.0830e-01   0.11   0.09   0.24 
Three Fastest
Protocols
1st2nd3rd
  e5   e4   c5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   7   31 

4x16 Processors (Column Major Ordering) / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.5834e-01   0.16   0.15   0.39 
Three Fastest
Protocols
1st2nd3rd
  e5   e4   e3 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   4   27 

2x32 Processors (Column Major Ordering) / Problem T42L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  6.7116e-01   0.26   0.25   0.65 
Three Fastest
Protocols
1st2nd3rd
  e3   c4   e5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   3   16 

8x8 Processors (Column Major Ordering) / Problem T85L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.0076e+00   0.05   0.05   0.12 
Three Fastest
Protocols
1st2nd3rd
  a1   a2   c4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   17   31 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:17:05 EDT.