Not anymore extreme than a bunch of ppl here accusing [H] of being AMD biased. Just because he games at 1600p or 3x1080p and thats where radeons perform better doesn't make him biased, it makes his results extremely valid for that market segment. Top cards for top monitor gaming.
Some of his results are biased because he pretends that 30-35 fps at 2560x1600 4AA is playable in a first person shooter or a racing game. If one card is doing 32 fps and another 25 fps, you can say that card 1 is faster, but both are completely unplayable. So really the fact that 1 is faster is irrelevant since you'll need to back off your image quality or lower the resolution. It's OK to benchmark BF3 and get 40 fps. But since his
entire benchmarking methodology is based on
his own interpretation of what playable is, and his standards are so low, and because his image quality settings are different for each videocard, it's impossible to say which card is actually faster.
Yet, in [H]'s world, 30-35 is perfectly fine for gaming. I have 0 problems with 2560x1600 or 3x 1080P benchmarks at real world 50-60 fps. If one card gets 50-60 and another 30-40, yes there is a winner. But comparing 25 vs. 32 fps or 37 vs. 42 fps and declaring a winner is like comparing a turtle to a snail in a race.
[H] needs to be reminded that PC gamers who play FPS and racing games prefer gaming on the PC over consoles because we like 60 fps, not 30-35 fps in those genres. But when all of your benchmarks are running at 1920x1080 8AA or 2560x1600 4AA at 25-40 fps, what am I suppose to extra from that, that my card isn't fast enough at those settings? I already know that.
[H] lost all credibility in my eyes when they said MLAA image quality was better than MSAA. Since then he uses FXAA and MLAA in all his reviews, making most of them worthless since FXAA and MLAA cause texture blurring. But that problem doesn't apparently exist for [H]. How convenient.
More so, I read Bit-Tech, TechReport, TechSpot, Xbitlabs, Digit Life, Computerbase, AnandTech, Hot Hordware, Legit Reviews, GameGpu.ru, Tom's Hardware and some others. Among a dozen of these professional reviews, [H] is the only website which
consistently shows HD6970 either as fast or beating GTX580. Every generation, he always comes up with benchmarks that somehow show AMD's lower card performing similar or better than NV's best card (and this has been a trend for at least 5 years+). I find it rather odd.