Why the 10-year plan is feasible...

[HappySqurriel]

Squilliam said:

HappySqurriel said:

Squilliam said:
In 2011 at 32nm process node a console manufacturer will be able to release a console which is at least 4x more powerful than the current generation of consoles on half as much silicon. Progress is always going to happen and people won't understand why such power is needed until it actually starts being used.

 

It is possible that by the end of 2011 (and certainly by the end of 2012) we could be working with a 22nm process which would mean that you could produce something 4 times as powerful as the PS3/XBox 360 at the cost to manufacture the Wii.

Consoles are usually produced on mature process nodes. Even Intels Tick/Tock cycle would put them as just implementing the 22nm process node at the end of 2012 and the other silicon manufacturers are behind Intel in this regard and the gap is more likely to increase rather than decrease as the shrinks become harder to implement.

The only console released in 2012 which would be likely incorperating chips produced at 22nm would be one which incorperates an Intel processor of some kind.

OT: Its funny, these pictures don't look any better than tech demos released by Nvidia or ATI as far back as 2005 and they were rendered in real time as well.

 

I'm just looking at it based on the 18 month cycle for manufacturing processes ...

IBM had the 90nm process for the XBox 360 at the end of 2005, which translates to a 65nm process in the middle of 2007, a 45nm process at the end of 2008, a 32nm process in the middle of 2010, and a 22nm process at the end of 2011. If a console is released at the end of 2011 and beyond it will (certainly) be using a 32nm process, but it is possible they may have a 22nm process; and even if they don't have the 22nm process it will (most likely) be possible within 12 months of releasing the system so they could assume they would get the cost savings from the process soon after they released the console.