sliderider wrote:It looks like you need a faster CPU. Those results are probably CPU bound. Either that, or the benchmarking program you used isn't sufficient to test those cards.
Think you're seeing it from the wrong perspective. Not trying to find the maximum speed these cards can go, just to see if there were any difference on a Pentium 100.
I want to create a Socket 5 & 7 motherboard/chipset benchmark comparison thread, and the Matrox cards are the best way to go about it for collecting results. Why Matrox? They are uniform. They didn't come in 10.000 different versions as the S3 Virge's, or other chipset which had several different manufactures. They are cheap and in good supply. Most members here on Vogons own one. By forcing Matrox cards from the Millennium II, Mystique and G200 groups you avoid graphic cards being a problem in the results.
Why Pentium 100? It is in good supply and works on all Socket 5 and Socket 7 boards. People can downclock their higher clocked Pentium's if they'd like (tested and makes no difference in results). It is also the comparison speed used by feipoa in his 486 benchmark.
From this table I found, G200 should be substantially faster.
It most likely is in Windows and 3D applications. But this is DOS VGA speed.