A few things to note here:
- The version of the 9800GT here is identical to the 8800GT. This card does not possess the die shrink of the G92. If it had their would have been 2 sli connectors as found on the cards with the smaller G92
- On the Uningine tests the 8800GT was slower because it was running out of video ram in these tests @ 2560 x 1600. I expect that Michael was using the 8800GT 256M that he had tested earlier. PTS supresses these messages unfortunately. Had the test been run straight from cli with out PTS the issue would have been readily apparent.
- Coolbits could very be well broken in 180 drivers. This might be a regression that was found in a earlier driver but was fixed with 177.80.
- There is no reason why a 8800GT with the same amount of ram would be any different in results at the same clock speed when compared to this particular card.


Reply With Quote
.

