Most important in my opinion are the texel/texture fill rate and pixel fill rate. Bandwidth can give you some hints how fast a GPU may be, but in most cases the bandwidth is chosen so high that the GPU is not fast enough to really use all of it. GFlops has the problem that GPUs in a gaming environment aren't doing the kind of things which are measured in GFlops, but atleast the number of flops can give you some hints how fast a GPU is (more than the bandwidth in my opinion).
But sometimes it can be difficult to get all those values (http://www.hwcompare.com/ has a lot but e.g. not the ones for HD 6550D).
And regarding the rumoured specs of the PS4: In the rumour I know of it was totally clear that the PS4 is using an A8-3850 and(!) a GPU based on the HD 6670.
I also think that it would not be possibly to put a GPU like the HD 6670 into an APU, because that would be really difficult to cool (but I'm no expert).
But to be clear: I for myself somewhat doubt the rumour because I'm not really sure if an asymmetric crossfire is really the way to go (while I'm still thinking it may work). There are some nice things you can do better if you have a GPU really near to your CPU but in this cases a smaller integrated GPU may as well do that jobs.
But if you put these rumours in your tables you have to list all important facts about the chips.
Existing User Log In
New User Registration
Register for a free account to gain full access to the VGChartz Network and join our thriving community.
Essay Pro