PSTSWM SP3-200 Communication Protocol Performance

Performance Studies using

PSTSWM

IBM SP3-200 Protocol Performance

(distributed FFT experiment A2)

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
Parallel Algorithm: distributed FFT
Number of Timesteps: 12
Results:

16x1 Processors / Problem T21L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.9103e-02   0.49   0.50   1.11 
Three Fastest
Protocols
1st2nd3rd
  c2   c1   c3 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   3   8 

32x1 Processors / Problem T42L1
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  7.3106e-02   0.36   0.32   0.66 
Three Fastest
Protocols
1st2nd3rd
  c4   a5   c5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   4   8 

8x1 Processors / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  2.6415e-01   0.25   0.27   0.64 
Three Fastest
Protocols
1st2nd3rd
  c5   c4   d4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   2   13 

16x1 Processors / Problem T85L1
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.9347e-01   0.20   0.19   0.42 
Three Fastest
Protocols
1st2nd3rd
  c5   c4   d4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   4   17 

32x1 Processors / Problem T85L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.8339e-01   0.25   0.28   0.61 
Three Fastest
Protocols
1st2nd3rd
  c5   c4   d4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   2   10 

8x1 Processors / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  2.3158e+00   0.16   0.13   0.60 
Three Fastest
Protocols
1st2nd3rd
  c5   c3   d4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   7   24 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:15:41 EDT.