Originally Posted by metala
You know that the bigger the number ,the bigger the performance. But what number? Shader processors, yes. But also the clock. And you don't know if there are fixes in the old r600 architecture to make it better in r700.
Dude, you keep spouting this ignorance everywhere I see ya. (no offence) but I hate repeating myself.
Sigh... Here goes again:
Less shader processors = less money to make = cheaper card money wise. But if you can optimize your bus, architecture and memory bandwidth to push as much as you can out it's rear end, then you can equal the performace of a higher card for less money, AND less power usuage. In the video card world, memory bandwidth is everything. This link may give you some idea:
Hope you spot the trend....
If your younger and don't get it, I'll give ya an analogy:
A Car with 8 cylinders and a Car with only 6 decide to race. Which do you think would win?
The one with the most horsepower and torque wins. Cylinders doesn't mean anything. Memory bandwidth is exactly that. Horsepower and torque. It's not something the videocard industry likes to talk about because for one, it's a weakness. People focus on 'oh the GPU is clocked at 3Ghz!! I'm gonna buy that FTW!' It's a selling point. Most (and obviously some here) have no clue what memory bandwidth is, or what it's used for... so why bother advertising it? Even my grandma knows how fast her CPU is though... THATS what their gonna focus on.
And thats exactly what the industry wants. Which is the stupidest thing someone could do. (without obviously checking the "horsepower" first)
Ignorance is bliss..