Originally posted by: coldpower27
Originally posted by: RussianSensation
Originally posted by: coldpower27
Originally posted by: RussianSensation
Update:
Xbitlabs review -
8600GTS loses in every single benchmark except Oblivion 1280x1024
It isn't even close. 8600GTS doesn't outperform X1950Pro, and gets absolutely pummeled by X1900XT.
They tested resolutions that 8600 GTS wasn't designed for like 16x12 and 19x12. It's expected 32 SP is much lower then expected with that low an amount of SP, not much can save it sort of games very favourable to this architecture.
Graphics cards are generally compared on features and price. Since there are no DX10 games out there and X1900XT's image quality is anything but lacking, considering 8600GTS costs $199-229, why shouldn't its performance be compared to cards in a similar price range? I don't buy a graphics card to "pimp" my rig out or for emotional satisfaction. I buy it based on performance for the $. If another card can play games at 1600x1200 at double the frames (i.e. X1900XT) why should I consider 8600GTS ?
If you recall Geforce 5600 which was quickly eclipsed by rev 2 and then by 5700Ultra. Perhaps 8600GTS will follow the same fate.
It should be compared to a card that has the same introduction MSRP, so the comparison is actually normalized. It doesn't matter if X1900 XT can beat this card regardless, this is a mainstream card, and it's designed to be cost effective, it's not realistic given it's transistor budget to expect it to beat the X1900 XT.
So it should be compared to the 7900 GS or X1950 Pro. A 7600 GT as well just to see how it compared to the first $199US unit of the Geforce 7 Series.
It's almost always the case where depreciated high end cards are better deals then new mainstream cards until the supplies right out of them because of the initial higher cost of the new mainstream variants.