I sent an email over to everybody in the news department at Gamespot regarding the article "TNT Blows Voodoo Away."
headline.gamespot.com
Hope they will check their figures. The only thing that I can think of to explain the numbers would be that they tested on DX 6, with unoptimized V2 Drivers. Also, the resolution of 8x6 may have something to do with it...are other game gauge scores done at 6x4? Also noted that the TNT board uses SGRAM. Doubt the final cheap street boards will use it...
The game gauge benchmark is a good idea, but the implementation is a bit weak. Problem: weighting. What if I set up a little stock index with a share of TDFX, STBI, CREAF and Berkshire Hathaway? Up a thousand, down a few thousand. The little guys don't matter much. On game Gauge, Q2 is clearly the most important number, yet it is overshadowed by just about all of the other games. I would have made this an issue via an email to CGW or ZD, but weighting a game that supports Single Pass Multi-texturing would be good for TNT and bad for Banshee :) So I will keep my mouth shut.
One other problem with the review: It subliminally suggests that TNT has better technology than TDFX. Though we know that it is "older" technology. We need the comparisons to be against Banshee!
Am going to suggest that 3Dnews.net do a survey regarding V2 and SLI... |
