that dose not show flexibility a new chip would, this shows that it can be castrated, cutting those clusters out dose not help power consumption so its clocked lower and will suck up extra power for no reason.
there is no chance of this beating the 5870m, the clock speed is to low and that low clock speed will make its texture processing way to slow like close to a mobile g92 but with more geometry like what a 260 original would have. there is no chance that it will compete with the 5870m. the 5870m is close to the 5770 so since the desktop 465 is slower than the desktop 5850 and the desktop 5770 isnt to far behind so with this massive clock down i would expect it to be bottleneck its self in the texture fill rate.
i will agree that its an accomplishment to get it sub 100W but i would expect that for the gpu not the platform.
we really need that new fermi die to bring the prices down, it should even match the stupidly giant die in performance with normal power and normal cooling
Bookmarks