PSTSWM AlphaSC-500 Communication Protocol Performance

Performance Studies using

PSTSWM

Compaq AlphaServer SC Protocol Performance

(transpose LT experiment II-A1 / O(log P) transpose algorithm)

Date/Person: November 4, 1999 / P. Worley
Platform: Compaq AlphaServer SC at Compaq (fsehpc):
     16 ES40 4-way SMP nodes (500 MHz Alpha 21264 with 4MB L2 cache)
Environment: Digital UNIX V5.0;   RMS 2.3.2
Code Version: 6.6.3
Compilation Options: f90 -O4 -assume accuracy_sensitive -math_library accurate -aling dcommons -align records -arch host -tune host
Math Library: none
Communication Library: MPI
Parallel Algorithm: logtrans
(assuming transpose FFT)
Number of Timesteps: 12
Results:

1x16 Processors / Problem T21L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  2.5077e-02   0.11   0.13   0.22 
Three Fastest
Protocols
1st2nd3rd
  a1   c2   a3 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   7   24 

1x8 Processors / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.1936e-01   0.06   0.06   0.11 
Three Fastest
Protocols
1st2nd3rd
  c3   c1   c5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   8   24 

1x16 Processors / Problem T85L1
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.8717e-01   0.03   0.02   0.06 
Three Fastest
Protocols
1st2nd3rd
  b2   b0   c2 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  7   19   24 

1x8 Processors / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.3421e+00   0.06   0.06   0.17 
Three Fastest
Protocols
1st2nd3rd
  a6   a2   b0 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   9   24 

Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:03:22 EDT.