PSTSWM SP3-375 Communication Protocol Performance

Performance Studies using

PSTSWM

IBM SP3-375 Winterhawk II Protocol Performance

(transpose LT experiment II-A / O(P) swap 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
Communication Library: MPI
Parallel Algorithm: swtrans
(assuming transpose FFT)
Number of Timesteps: 12
Results:

1x16 Processors / Problem T21L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  2.2160e-02   1.52   1.53   2.97 
Three Fastest
Protocols
1st2nd3rd
  e0   e1   e2 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   3   4 

1x32 Processors / Problem T42L1
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.3072e-02   2.31   2.41   4.22 
Three Fastest
Protocols
1st2nd3rd
  e0   e2   e1 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   1   4 

1x8 Processors / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  8.2395e-02   0.32   0.32   0.59 
Three Fastest
Protocols
1st2nd3rd
  e3   e5   e4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   3   11 

1x16 Processors / Problem T85L1
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.2131e-01   0.43   0.40   0.81 
Three Fastest
Protocols
1st2nd3rd
  e4   e5   e3 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   3   7 

1x32 Processors / Problem T85L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.5838e-01   0.69   0.67   1.35 
Three Fastest
Protocols
1st2nd3rd
  e4   e5   e3 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   3   4 

1x8 Processors / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  7.7999e-01   0.12   0.11   0.65 
Three Fastest
Protocols
1st2nd3rd
  e5   e3   e4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   6   30 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:19:52 EDT.