The price/performance leader is clear, though: this honor goes to the BFG GeForce GTX 275. For those in search of high performance but who don't have unlimited budgets, this card does the job. We also give an Editor's Choice to the MSI GeForce GTX 285 SuperPipe because of its great balance between performance and cooling, which shows that even an air-cooled high-end graphics card can still be very quiet.
That review failed hard with the drivers, settings, games, cards tested (no 4890? how can you say the GTX 275 is the best price/performance wise?) not to mention that the conclusion was ridiculous, "the fastest cards need watercooling"?! No they don't, my 2 4870 X2s don't need watercooling in a case that performs less than Antec 900 in air cooling when it is using the stock configuration. Yes they are watercooled but that is more for the enjoyment from watercooling for me, not necessity or any from of practicality.
Worst roudnup I have seen in a long long time. Where were all the ATI cards? How can you draw conclusions without even comparing the 4890's.. I was really interested to see the 4890's at 1Ghz vs an overclocked 275 or 285. What is with the games chosen too? Why on earth include 3dmark 06 and not vantage? Why use HAWX if you are not going to use DX10.1? The game sucks Did they even use the correct drivers? My 4890's perform substantially better than that.. I don't have an i7 either..
Bad article, even when compared to the crap Tom's has been putting out lately.
Tom's has actually been quite decent lately. I have to say though, when I looked at the benchmarks I was really confused. Old games, no Vantage or Crysis, no World in Conflict? 3D mark 06 is three years old, now! We're using high end cards and the tests are performed at 1680x1050 and 1920x1200? Sure, 1920x1200 is big, but 2560x1600 would really be the top-tier test. I think, by some of the results almost being >100 fps across the board, it's clearly stated that higher resolutions are needed...or more demanding games. I don't know why they threw in the blurb about the triple-decker 4870x2 and then only actually test the reference card. (maybe I just missed an explanation in the reading, though). The conclusion is fine, as long as the reader is only considering nvidia cards - it doesn't draw a good match to ati's price points. I just don't understand some of the selective testing.
Agree that this review could use improvement in finding more rigorous benchmark methods rather than "oh these cards does that 125 frames per sec and that does 130 fps."
But maybe there's a point to be made that the enthusiast who will cough up a small fortune for the best GPUs will also need the best CPUs and mobos and 30" monitors to reap the full potential of these high end cards.