Performance Studies using PSTSWM

Intel Paragon Protocol Performance

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

1x16 Processors / Problem T10L16

Runtime Statistics min (mean-min)/min (median-min)/min (max-min)/min   3.6897e-01    0.09    0.09    0.20

Three Fastest Protocols 1st 2nd 3rd   e2    e3    c2         Number of Proctocols With Runtimes Within X% of Min 1% 5% 25%   2    4    29

1x8 Processors / Problem T21L8

Runtime Statistics min (mean-min)/min (median-min)/min (max-min)/min   1.0409e+00    0.02    0.01    0.05

Three Fastest Protocols 1st 2nd 3rd   d2    d3    b6         Number of Proctocols With Runtimes Within X% of Min 1% 5% 25%   11    28    29

1x16 Processors / Problem T21L16

Runtime Statistics min (mean-min)/min (median-min)/min (max-min)/min   1.1336e+00    0.02    0.01    0.06

Three Fastest Protocols 1st 2nd 3rd   a2    d2    a3         Number of Proctocols With Runtimes Within X% of Min 1% 5% 25%   10    28    29

1x32 Processors / Problem T21L32

Runtime Statistics min (mean-min)/min (median-min)/min (max-min)/min   1.3369e+00    0.04    0.03    0.11

Three Fastest Protocols 1st 2nd 3rd   b0    a2    a3         Number of Proctocols With Runtimes Within X% of Min 1% 5% 25%   5    21    29

1x8 Processors / Problem T42L16

Runtime Statistics min (mean-min)/min (median-min)/min (max-min)/min   9.7197e+00    0.02    0.02    0.06

Three Fastest Protocols 1st 2nd 3rd   d6    b4    b5         Number of Proctocols With Runtimes Within X% of Min 1% 5% 25%   10    28    29

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 10:30:15 EDT.