Next: CONCLUSIONS
Up: Implicit finite-difference migration
Previous: Compiler notes
Figures 9 and 10 show the
finite-difference benchmark results for the four platforms.
As in the previous tests there was little separating the MIPS-based
quad processor SGI with the Intel Xeon-based quad processor Linux.
The results appear to be similar to the larger split-step migration
benchmark.
fdmig.par
Figure 9
Relative speed of the parallel portion
of the finite-difference migration on the various testing platforms.
1 represents the SGI, Origin 200; 2 is the
SGI 1400L ; 3 is the VA Start X MP; and 4 is the SGI
Power Challenge. In each case the solid line represent actual
performance, the dashed lines ideal performance.
fdmig.tot
Figure 10
Relative speed of the total execution time of
the finite-difference migration on the various testing platforms.
1 represents the SGI, Origin 200; 2 is the
SGI 1400L ; 3 is the VA Start X MP; and 4 is the SGI
Power Challenge. In each case the solid line represent actual
performance, the dashed lines
ideal performance.
Next: CONCLUSIONS
Up: Implicit finite-difference migration
Previous: Compiler notes
Stanford Exploration Project
10/25/1999