Xbox.com | Community - GamerDads 2.0 (aka Gamerparents 2.0)

Gamer Parents & Family Gaming

Started by xenocidic at 02-16-2008 8:38 PM. Topic has 6119 replies.

Sort Posts:

Page 202 of 204 (6,120 items)

« First ... < 200 201 202 203 204 >

02-16-2008, 8:38 PM

M

Re: GamerDads 2.0 (aka Gamerparents 2.0)

Unreal Butcher wrote:

Re-designing a CPU core to a smaller die-area helps achieve several of these goals.

Shrinking everything (a "photomask shrink"), resulting in the same number of transistors on a smaller die, improves performance (smaller transistors switch faster), reduces power (smaller wires have less parasitic capacitance) and reduces cost (more CPUs fit on the same wafer of silicon).
Releasing a CPU on the same size die, but with a smaller CPU core, keeps the cost about the same but allows higher levels of integration within one VLSI chip (additional cache, multiple CPUs, or other components), improving performance and reducing overall system cost.

Pretty much the things in red above. When they do a die shrink it usually means it'll use less power to run that particular part, which in turn makes it run cooler. Sounds like one of them would have the CPU and GPU on one die similar to the APX 2500. Which should mean a savings in power, cheaper manufacturing costs (as I wouldn't think you'd need seperate manufacturing lines for the CPU and GPU and you're using less material) and a cooler running 360.

Remember my above post is just from what I understand, and I could be completely wrong. If anyone has a better understanding, by all means step up and edgeumacate us. Thanks.

OK, as someone who works for the company and in the department that will actually help build this item, I think I'm certified to confirm all of the above as correct. Valhalla WILL be a CPU/GPU integrated processor. It also will be built off the same technology we use to build supercomputer processors. Now, granted, smaller die=more transistors=less wafer cost=low power...but you still have challenges at smaller dimensions with leakage and cross-talk between the devices. Also note that the size "savings" is most likely getting eaten up by plopping down a monster of a GPU. So you have to do fancier and fancier things with the process to reduce these risks along with knowing how best to integrate GPU and CPU on the same chip without eating up more silicon than you bargained for. Also note that lower power doesn't ALWAYS mean "cooler". More "turned on devices" on a smaller footprint could result in some runaway thermal effects that need to be managed by more efficient packaging and heatsinks. These babies go through all the motions with regard to this "full on" effect to see how long they last before they ship. So you can bet this next chip gets the same treatment...but with a GPU with transistor logic going and CPU with all its logic firing away as well at the same time on the same eentsie bit of silicon...expect the heat to come with it.

Hope that wasn't too painful of anedeumacation.

Edit: oh, please note, new game in 'da house this weekend (OrangeBox). My excellent work on Valentine's Day appears to have paid off.

This posting is provided "as-is" and confers no rights

Report

02-17-2008, 4:01 AM