Full PlayStation 4 specifications

[joeorc]

goddog said:

the rummored 1.6 would not cut it. go back and look at the IPC of the liniage there yes it has imporved but it needs to be about 15-20% higher clock than comparable intel chips to get close to same leve of output (depeding on work type) even with "8 cores" you will have issues with some games playing at 1.6  . go ahead try out cpu bound games like skrim or civilization, while it will not hurt all games, it will lead to hampering the gpu and limiting the ability to use all resources (throttle a cpu back and play them its not hard to do) and in some cases make them unplayable. 

 

now if its an idle of 1.6 with throttling under load of up to say 4.x  depending on thermals that would work, but could lead to inconsistant game play among consoles 

more practically 1.6 base with a boost to 3.x  designed to keep it form any chance of throttling would work but would agian later in the gen lead to the gpu being hindered 

 

again so this is all not negative props on the gddr5 8 gig did not expect that, and it will be very interesting to see how that ends up being used and how it helps the cpu. I had been expecting at best 4 gig ddr 3 for the cpu and 2-3 gig gddr5 for the gpu , but realistically  4 gig ddr3 for the cpu and 1 gig gddr5  they blew that away and bravo to them

it would be nice to know what gpu family line the custom chip came from  tahiti or pitcarin 

upon reflection of the gpu id like to know the break down of how it was modified/clustered a bit more... it could be more powerful than we think for gaming with some of its gpucompute removed if the right optimizations were made, AMDs 7xxx series chips have left  some gaming on the table to ramp up the gpucompute and costimization could boost performance drastically . could be up to a 7890 performanc  (should have been called a 7930 to avoid chip confusion but both gpu makes seem to hate customers understanding what they are buying )

 

while what your saying may have a bearing, and like you stated we do not know how its set yet in the design, what i do know is if it's anything like they did with the PSVita's 3D TSV i do not think the process will be bottlenecked due to the fact that since the GPU is in the same Die as the CPU and the Ram is also a 3D stacked TSV inside the same SOC package connected via a powerful High speed wide Bus there really would not be a real bottle neck to offload the problem to say a n. or S. bridge like you Have in some designs because every connection is less stepping cycles required.