PSTSWM Paragon Communication Protocol Performance

Performance Studies using

PSTSWM

Intel Paragon Protocol Performance

(distributed FFT experiment A1)

Date/Person: May 13, 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: KAI
Communication Library: NX
Parallel Algorithm: distributed FFT
Partition: 1x8, 1x16, or 1x32
Results:

16x1 Processors / Problem T21L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  3.2902e-01   0.06   0.05   0.19 
Three Fastest
Protocols
1st2nd3rd
  a2   a1   a3 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  4   16   24 

32x1 Processors / Problem T42L1
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  4.3844e-01   0.04   0.03   0.15 
Three Fastest
Protocols
1st2nd3rd
  c2   d0   c3 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  3   17   24 

8x1 Processors / Problem T42L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.7431e+00   0.05   0.05   0.11 
Three Fastest
Protocols
1st2nd3rd
  b4   b6   b5 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  7   12   24 

16x1 Processors / Problem T85L1
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  2.4960e+00   0.05   0.04   0.13 
Three Fastest
Protocols
1st2nd3rd
  b4   b3   b2 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  7   14   24 

32x1 Processors / Problem T85L2
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  2.9660e+00   0.07   0.07   0.19 
Three Fastest
Protocols
1st2nd3rd
  b3   b6   b0 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  7   7   24 

8x1 Processors / Problem T85L4
Runtime Statistics
min(mean-min)/min(median-min)/min(max-min)/min
  1.8116e+01   0.03   0.03   0.08 
Three Fastest
Protocols
1st2nd3rd
  b5   b4   b6 
       Number of Proctocols With
Runtimes Within X% of Min
1%5%25%
  4   17   23 

DISCUSSION

Partitions of the Paragon processor grid were used that match the processor subsets that the parallel algorithms would run on in a two dimensional data decomposition: 1x8, 1x16, and 1x32.
Patrick H. Worley / ( worleyph@ornl.gov)
Last Modified Monday, 15-Jul-2002 10:29:27 EDT.