dcIKeeL said:
More importantly, I never said the 360 slim was coming out this year. 360 price cuts are coming this year but the slim is going to come as a strategic slap in the face sometime in mid 2010. Perhaps MS will even make the slim coincide with Natal. Infamous8 was claiming that MS isn't going for a true slim because they already have their 'slim' with the arcade. However the arcade has received modifications that make it even cheaper to manufacturer (if only a bit) so the arcade itself has been 'slimmed' down and will see price cuts. MS true slim model is still on it's way. Sony I'm pretty sure is releasing their ps3 slim this fall. Investors think of these rumors as more than just rumors as of now since they are willing to invest on it. It will be 120gb slim ps3. http://seekingalpha.com/article/146289-why-sony-s-ps3-will-continue-to-struggle?source=yahoo |
I think you're missing the most important point. A cutting edge fab process is not the most important consideration when it comes down to producing console chips. They can only benefit from half the equation which is more dies per wafer and lower power consumption. They cannot increase the clock speeds or implement new technology so at this point what they want is a stable and mature fab process with good yields. It costs more to use a cutting edge fabrication node and the yields can be worse so the cost of going to 32nm may very well be more than the cost of sticking to 45nm. Just because they can use a node, doesn't mean they will and so far they have been very conservative so why would they change now?
They have several intermediate steps to consider first:
- Combine Xenos + ED Ram on the same die, 55nm is mature and as its a half node they can use that.
- Change the ram technology to GDDR5 and use a slimmer 64bit bus and the 2Gbit chips, so 4 ram chips vs 8, and eventually 4Gbit chips to use only 2 as the price of this technology comes down and the price of the GDDR3 goes up. Its probably the biggest cost cutter they could implement this year.
- Shrink the CPU down to 45nm, the process is mature so why not and that paves the way for combining the chips @45nm next year on the same process.
Tease.







