PSTSWM SP3-375 Communication Protocol Performance

Performance Studies using

PSTSWM

IBM SP3-375 Winterhawk II Protocol Performance

(distributed LT experiment C / exchange sum 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
Parallel Algorithm: exchsum
Number of Timesteps: 12
Results:

4x16 Processors (Column Major Ordering) / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.3660e-01   0.03   0.03   0.06 
Three Fastest
Protocols
1st2nd3rd
  c2   a5   d4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   23   25 

2x32 Processors (Column Major Ordering) / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.4243e-01   0.10   0.11   0.20 
Three Fastest
Protocols
1st2nd3rd
  a4   c1   a0 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   6   25 

8x8 Processors (Column Major Ordering) / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.8150e-01   0.02   0.02   0.06 
Three Fastest
Protocols
1st2nd3rd
  c4   d5   d4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   24   25 

4x16 Processors (Column Major Ordering) / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  6.2382e-01   0.05   0.03   0.13 
Three Fastest
Protocols
1st2nd3rd
  b0   b1   b2 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  5   17   25 

2x32 Processors (Column Major Ordering) / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  9.2949e-01   0.12   0.06   0.35 
Three Fastest
Protocols
1st2nd3rd
  a1   a4   a5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   12   19 

8x8 Processors (Column Major Ordering) / Problem T85L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.2413e+00   0.05   0.06   0.09 
Three Fastest
Protocols
1st2nd3rd
  a1   a5   b0 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   11   25 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:19:14 EDT.