Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: niceguy767 who wrote (229919)	4/14/2007 1:16:42 PM
From: neolib	Read Replies (1) \| Respond to of 275872

This weeks EETimes has a front page article on foundries push to 45nm for late this year. TSMC is pulling in their's a bit (Sept for first customer wafers instead of Q4), with UMC and IBM/Chartered/Samsung shortly behind by years end. However, a scary stat listed in another article (p60) makes the following claim: At 65nm an IDM needs $8.3B/yr in sales to be profitable. At 45nm the estimate is $13.3B/yr At 32nm the estimate is $16.7B/yr. This is looking at the cost to an IDM for the entire process node, so it is independent of #/fabs (at least my reading). To what degree AMD can lower this bar by tying in with IBM/Chartered I'm not sure. Clearly AMD is not hitting the 65nm metric, and the odds of them growing to the 45nm level in the time required is zero.

To: niceguy767 who wrote (229919)	4/14/2007 5:47:32 PM
From: pgerassi	Read Replies (1) \| Respond to of 275872

Dear Niceguy767: Intel won't get that much an advancement in capacity because they will use double exposure. That causes the 45nm steps to multiply by 7 for an overall step count of about 50% more and thus a yield loss of another 20-30%. multiplying these factors with the dice size reduction (assuming no cache or core transistor count increases) and z Fab making X 65nm CPUs now will make somewhat less than X 45nm CPUs. So Intel is spending all that capex on 45nm and getting no more capacity. Their per die costs will jump with all of the added depreciation. AMD with immersion will increase its capacity by 80-120%. Even with the capex reduction, that will only go down to 70-100% increase with lower per die costs. Intel did it to cut two quarters from 45nm production start. Evidently they thought that if they went with immersion, AMD would be one quarter or less behind them and they know they don't compete well with AMD on the same process node. They figured if that happened, they would not ever get the performance crown back. Pete PS: This double exposure penalty is being completely ignored by WS and analysts. When it becomes evident to them, watch Intel stock price fall off a cliff.