PSTSWM SP3-200 Communication Protocol Performance

Performance Studies using

PSTSWM

IBM SP3-200 Protocol Performance

(distributed LT experiment B2 / recursive halving sum algorithm)

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: halfsum
Number of Timesteps: 12
Results:

4x16 Processors (Row Major Ordering) / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.0493e-01   0.18   0.20   0.34 
Three Fastest
Protocols
1st2nd3rd
  c5   c4   d4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   2   16 

2x32 Processors (Row Major Ordering) / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  9.1384e-02   0.31   0.23   0.60 
Three Fastest
Protocols
1st2nd3rd
  c4   c5   a4 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   1   13 

8x8 Processors (Row Major Ordering) / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.6726e-01   0.06   0.05   0.22 
Three Fastest
Protocols
1st2nd3rd
  a5   c5   c1 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   15   25 

4x16 Processors (Row Major Ordering) / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.7673e-01   0.09   0.09   0.20 
Three Fastest
Protocols
1st2nd3rd
  a4   a5   c1 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   6   25 

2x32 Processors (Row Major Ordering) / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  5.5596e-01   0.17   0.18   0.30 
Three Fastest
Protocols
1st2nd3rd
  a5   a4   a1 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   2   20 

8x8 Processors (Row Major Ordering) / Problem T85L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.3918e+00   0.05   0.04   0.12 
Three Fastest
Protocols
1st2nd3rd
  c1   a4   a5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   15   25 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:15:16 EDT.