PSTSWM SP3-375 Communication Protocol Performance

Performance Studies using

PSTSWM

IBM SP3-375 Winterhawk II Protocol Performance

(distributed LT experiment C / recursive halving 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: halfsum
Number of Timesteps: 12
Results:

4x16 Processors (Column Major Ordering) / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.1298e-01   0.09   0.08   0.16 
Three Fastest
Protocols
1st2nd3rd
  c5   c4   d5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   5   25 

2x32 Processors (Column Major Ordering) / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  7.0393e-02   0.25   0.21   0.47 
Three Fastest
Protocols
1st2nd3rd
  c4   c5   d4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   2   14 

8x8 Processors (Column Major Ordering) / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.3761e-01   0.03   0.02   0.05 
Three Fastest
Protocols
1st2nd3rd
  c5   c4   c2 
       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
  3.8401e-01   0.04   0.04   0.06 
Three Fastest
Protocols
1st2nd3rd
  c4   c5   d5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   20   25 

2x32 Processors (Column Major Ordering) / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.5425e-01   0.05   0.06   0.10 
Three Fastest
Protocols
1st2nd3rd
  c4   d5   b4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   9   25 

8x8 Processors (Column Major Ordering) / Problem T85L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  2.8357e+00   0.02   0.02   0.04 
Three Fastest
Protocols
1st2nd3rd
  b5   c5   b6 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   25   25 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:19:18 EDT.