Performance Studies using PSTSWM

Intel Paragon Protocol Performance

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

1x16 Processors / Problem T10L16

Runtime Statistics min (mean-min)/min (median-min)/min (max-min)/min   4.3829e-01    0.03    0.02    0.31

Three Fastest Protocols 1st 2nd 3rd   d2    b0    b6         Number of Proctocols With Runtimes Within X% of Min 1% 5% 25%   7    24    24

1x8 Processors / Problem T21L8

Runtime Statistics min (mean-min)/min (median-min)/min (max-min)/min   1.1418e+00    0.04    0.02    0.34

Three Fastest Protocols 1st 2nd 3rd   d2    b6    d6         Number of Proctocols With Runtimes Within X% of Min 1% 5% 25%   9    22    24

1x16 Processors / Problem T21L16

Runtime Statistics min (mean-min)/min (median-min)/min (max-min)/min   1.3036e+00    0.05    0.02    0.49

Three Fastest Protocols 1st 2nd 3rd   d2    b0    d4         Number of Proctocols With Runtimes Within X% of Min 1% 5% 25%   1    15    24

1x32 Processors / Problem T21L32

Runtime Statistics min (mean-min)/min (median-min)/min (max-min)/min   1.6216e+00    0.07    0.03    0.67

Three Fastest Protocols 1st 2nd 3rd   d6    b6    d3         Number of Proctocols With Runtimes Within X% of Min 1% 5% 25%   8    13    24

1x8 Processors / Problem T42L16

Runtime Statistics min (mean-min)/min (median-min)/min (max-min)/min   1.0360e+01    0.03    0.01    0.28

Three Fastest Protocols 1st 2nd 3rd   d4    d6    b6         Number of Proctocols With Runtimes Within X% of Min 1% 5% 25%   8    22    24

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 10:29:40 EDT.