PSTSWM SP3-200 Communication Protocol Performance

Performance Studies using

PSTSWM

IBM SP3-200 Protocol Performance

(distributed FFT experiment B2)

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:

16x4 Processors (Row Major Ordering) / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  9.5481e-02   0.28   0.28   0.46 
Three Fastest
Protocols
1st2nd3rd
  a5   c4   c5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   3   10 

32x2 Processors (Row Major Ordering) / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  8.4497e-02   0.37   0.44   0.71 
Three Fastest
Protocols
1st2nd3rd
  a4   a5   c4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   3   9 

8x8 Processors (Row Major Ordering) / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.7981e-01   0.14   0.15   0.32 
Three Fastest
Protocols
1st2nd3rd
  d5   a1   a4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   3   25 

16x4 Processors (Row Major Ordering) / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.6710e-01   0.14   0.15   0.27 
Three Fastest
Protocols
1st2nd3rd
  c5   d5   a4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   5   24 

32x2 Processors (Row Major Ordering) / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.2471e-01   0.25   0.26   0.46 
Three Fastest
Protocols
1st2nd3rd
  c5   a4   a5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   1   12 

8x8 Processors (Row Major Ordering) / Problem T85L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.6103e+00   0.11   0.12   0.28 
Three Fastest
Protocols
1st2nd3rd
  a5   d5   d4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  4   6   25 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:14:50 EDT.