| Spec | System A | System B | System C | System D |
| OS: | Gentoo, Linux 2.6.16-rc5 | Gentoo, Linux 2.6.8.1 | Mac OS X 10.4.6 | Gentoo, Linux 2.6.17-gentoo-r7 |
| Processor: | AMD64X2 +4200 | Pentium 4 HT | Pentium T2500 | Core 2 Duo EE6600 |
| Frequency (MHz): | 2200 | 2400 | 2000 | 2400 |
| L2 Cache: | 512KB x2 | 512KB | 2MB | 4MB |
| FSB (MHz): | 1000 | 800 | 667 | 1066 |
| Memory: | 2GB | 512MB | 1GB | 2GB |
| G++ opts: | -O3 -march=k8 | -O3 -march=pentium4 | -O3 -march=pentium-m | -O3 -march=pentium-m |
|
The benchmarking program is a Monte Carlo simulation. The simulation specs were chosen so as to avoid any memory size issues between the systems. The algorithm basically does a huge number of iterations on elements of a matrix (elements are of C++ type double). Each iteration preforms simple arithmetic on the matrix elements. The average of three trials are shown below.
|
