Originally posted by: hans007
Originally posted by: ilkhan
actually the recession allows for vastly cheaper capital costs for the next node, assuming the company can afford it. The suppliers are hurting and will take less to get any orders. Thus the 7B or whatever that intel just dropped on 32nm.
And yeah, we weren't expecting 32nm until november anyway. clarkdale/arrandale in november, gulftown in Q1, and the xeon gulftown in Q2 (probably).
not exactly.
most of these companies still have to evaluate what the best use of the money is.
Given the credit situation, the ROI of a new process will not be as much even if costs less to implement.
Not to mention given that money is fairly tight they could have better uses for that money (seeing as intel is in the lead they really have no reason to push farther ahead as fast).
I mean if you were intel and you were kicking amd's ass, would you spend money to push farther ahead for no real advantage with possibily a very underutilized factory in the future (seeing as you would be making way more 32nm chips when demand is falling , given there is no demand for faster stuff right now).
Or would you be better off taking your cash horde, and buying other companies / moving into new fields, loaning the money out at high interest rate since the credit market is tight for corporate bonds.
Just having a giant pile of money doesn't mean they are going to build a factory as it might not be the best use of it. Their shareholders would wnat them to make the most fiscally prudent choice and I think that choice is... sell more 45nm stuff and ramp 32nm slowly. I'm sure given enough money they could push it out faster, but as i've said theres better uses for that money.