Seems like RV770 has ROPs on steroids - AA is better than Nvidia now
That is an impressive score no doubt about that...
I am especially curious about the scaling improvements and what they did to actually improve things and what was wrong with the old design. CF scaling on the HD3x series was the main issue that made me not buy ATI products...
Aparently the bridge chip is gone. So they are using something else this time...
How about we see some REAL benchmarks before making any assumptions shall we! Isn't that what you said about Nvidia's 'benchmarks'??
So, yeah, you're always right.
Actually, games benchmarks are the ONLY benchmark that counts, as has been proven over the years. And as far as performance fluctuations based on personal experience, that's crap, no system changes enough to be any consequence to overall performance of a graphics card!
Like I have said before, performance is what matters to me, not the brand. If AMD beat Nvidia, then that is cool, but until they show benchmarks with real games and not synthetic benches, then, and only then will I consider an AMD (ATI).
While it's true that game benchmarks are important. Standard benchmarks have their use too. After all, it's a real benchmark! It accurately predicts the power of the hardware in its own way.
It's a nice score, but says nothing conclusive yet
It's FAKE!!! Look at bandwidth 57 GB/s It's not real, GDDR 5 has more... !!! I Think it's FAKE FAKE FAKE!
Also those images show the 4870x2 having 900 shaders and a 256-bit bus, now is it just me or should the card have 1600 shaders and a 512bit bus?
Core i7 940 @ 4.01Ghz - 6GB DDR3 1608Mhz -
SLI GTX 280 SSC - EVGA X58 Classified -
ASUS Xonar Essence STX - Corsair 1000W PSU