PSTSWM Paragon Communication Protocol Performance

Performance Studies using

PSTSWM

Intel Paragon Protocol Performance

(transpose FFT experiment A2 / O(P) sendrecv transpose algorithm)

Date/Person: June 4, 1998 / P. Worley
Platform: Intel Paragon XP/S 150 MP at Oak Ridge National Laboratory:
     1024 MP nodes (3 50-MHz iPSC/860 processors per node)
Environment: Paragon OSF/1 Release 1.0.4 Server 1.4 R1_4_5
f77/Paragon Paragon Version R5.0.3
Code Version: 6.3
Compilation Options: if77 -O4 -Mnodepchk -Knoieee -Msafealloc
Math Library: none
Communication Library: NX
Parallel Algorithm: srtrans
Partition: 4x2, 4x4, or 8x4
Results:

16x1 Processors / Problem T10L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  5.1294e-01   0.09   0.08   0.20 
Three Fastest
Protocols
1st2nd3rd
  e3   e2   c2 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  2   5   29 

32x1 Processors / Problem T10L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  6.2795e-01   0.13   0.09   0.33 
Three Fastest
Protocols
1st2nd3rd
  e3   e2   c2 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  1   4   27 

8x1 Processors / Problem T21L8
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.6390e+00   0.02   0.02   0.04 
Three Fastest
Protocols
1st2nd3rd
  d2   c2   d3 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  6   29   29 

16x1 Processors / Problem T21L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.7664e+00   0.02   0.02   0.04 
Three Fastest
Protocols
1st2nd3rd
  c2   c3   e3 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  9   29   29 

32x1 Processors / Problem T21L32
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  2.0052e+00   0.04   0.04   0.09 
Three Fastest
Protocols
1st2nd3rd
  c2   e3   e2 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  4   21   29 

8x1 Processors / Problem T42L16
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.5329e+01   0.02   0.02   0.04 
Three Fastest
Protocols
1st2nd3rd
  d6   b6   b5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  6   29   29 

DISCUSSION

The Paragon processor grid partitions were chosen to match those used for the October, 1994 SUNMOS data.
Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:30:06 EDT.