First post, by paradigital
So, after the saga that was getting my Slot A build stable (when caps go bad), I started trying a few video cards in the machine to get a feel for what felt best/provided the best performance whilst not being too far out in terms of period. To that end I wanted to try the following:
TNT2 Ultra (Diamond Viper V770U)
Geforce 256 DDR (Creative Labs Annihilator Pro)
Geforce 2 MX (Dell OEM, not M64, 200 or 400)
Geforce 2 GTS Pro (Dell OEM)
3Dfx Voodoo 3 3000
I started with the TNT2 Ultra, which turned in a 3DMark99 Max score of 6,843.
Before diving too deep into testing a suite of games I wanted to just compare 3DM99 to whittle out the slowest, so removed all video drivers, popped the MX in and installed drivers, rebooted and ran 3DM99.
The MX turned in 5,551. Seems a bit… low.
The CPU scores between the two cards are similar enough that I think I can rule out the influence of the graphics card on the mainboard/RAM/CPU speed, with the MX scoring 14,710 and the TNT2 14,795.
The MX then proceeds to spank the TNT2 in fill rate, texturing, rasterizing, bump mapping, indeed almost every synthetic benchmark was a clear win for the MX, anywhere from 20% increase (e.g. single-pass bump mapping) to 2000% increase in single point lighting.
However, in the trilinear texturing test and the two “game” tests, the TNT2 pulls ahead. A 4% win for the TNT2 in trilinear texturing, a 15% lead in “game 1” and a 23% lead in “game 2” for the TNT2.
Something feels amiss here, the TNT2 shouldn’t be pulling off a defeat of the MX, right?