PSTSWM T3E-900 Communication Protocol Performance

Performance Studies using

PSTSWM


SGI/Cray Research T3E-900 Protocol Performance

(transpose FFT experiment B / O(P) sendrecv transpose 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: MPI
Parallel Algorithm: srtrans
Number of Timesteps: 12
Results:

16x4 Processors (Row Major Ordering) / Problem T21L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.1929e-01   0.17   0.14   0.48 
Three Fastest
Protocols
1st2nd3rd
  a6   a0   e0 
      Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   2   25 

32x2 Processors (Row Major Ordering) / Problem T21L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  2.1608e-01   0.14   0.11   0.44 
Three Fastest
Protocols
1st2nd3rd
  e0   a0   e2 
      Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  4   8   25 

8x8 Processors (Row Major Ordering) / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.6611e-01   0.04   0.05   0.11 
Three Fastest
Protocols
1st2nd3rd
  e0   e2   c2 
      Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  6   16   31 

16x4 Processors (Row Major Ordering) / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.2545e-01   0.10   0.09   0.20 
Three Fastest
Protocols
1st2nd3rd
  e0   a0   e1 
      Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   8   31 

32x2 Processors (Row Major Ordering) / Problem T42L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  7.0759e-01   0.07   0.07   0.19 
Three Fastest
Protocols
1st2nd3rd
  c1   e0   c2 
      Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  6   11   31 

8x8 Processors (Row Major Ordering) / Problem T85L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.1173e+00   0.02   0.01   0.04 
Three Fastest
Protocols
1st2nd3rd
  c1   a0   c2 
      Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  14   31   31 


Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:28:00 EDT.