(on a single node)
Benchmark |
PPro 200 |
PII 300 |
Convex/HP |
Rolled Single Precision (spr) |
45.6 Mflops |
84.7 Mflops |
32.9 Mflops |
Unrolled Single Precision (spu) |
62.4 Mflops |
83.7 Mflops |
32.7 Mflops |
|
43.2 Mflops |
64.8 Mflops |
77.9 Mflops |
Unrolled Double Precision (dpu) |
50.1 Mflops |
66.0 Mflops |
77.1 Mflops |
Standard Linpack benchmark by J. Dongarra, ORNL – obtained from netlib
100x100 real matrix multiply
Herbert W. Hamber 10/24/97