PSTSWM T3E-900 Communication Protocol Performance

Performance Studies using

PSTSWM

SGI/Cray Research T3E-900 Protocol Performance

(distributed LT experiment B / recursive halving sum algorithm)

Date/Person: May 11, 1998 / P. Worley
Platform: T3E-900 at National Energy Research Scientific Computing Center (mcurie.nersc.gov):
   532 450-MHz DEC Alpha EV5 RISC processors
Environment: UNICOS/mk 2.0.2.28
mpt 1.2.0.2
Cray CF90 Version 3.0.2.1
Code Version: 6.3
Compilation Options: -dp -Oscalar3 -Wl,-Dstreams=on
Math Library: libsci
Communication Library: SHMEM
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
  4.1795e-02   0.05   0.06   0.11 
Three Fastest
Protocols
1st2nd3rd
  a2   a1   c2 
      Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   4   10 

2x32 Processors (Row Major Ordering) / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.6013e-02   0.07   0.11   0.14 
Three Fastest
Protocols
1st2nd3rd
  a1   c2   a2 
      Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   3   10 

8x8 Processors (Row Major Ordering) / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.4439e-01   0.02   0.02   0.04 
Three Fastest
Protocols
1st2nd3rd
  a1   c2   a2 
      Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   10   10 

4x16 Processors (Row Major Ordering) / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.8133e-01   0.03   0.04   0.09 
Three Fastest
Protocols
1st2nd3rd
  a2   a1   c2 
      Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   8   10 

2x32 Processors (Row Major Ordering) / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.2776e-01   0.05   0.05   0.15 
Three Fastest
Protocols
1st2nd3rd
  a1   a2   c2 
      Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   4   10 

8x8 Processors (Row Major Ordering) / Problem T85L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.6155e+00   0.01   0.01   0.04 
Three Fastest
Protocols
1st2nd3rd
  a2   c2   a1 
      Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   10   10 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:24:55 EDT.