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 : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: fingolfen who wrote (141478)	8/13/2001 4:28:43 PM
From: AK2004	Read Replies (1) \| Respond to of 186894

fingolfen re: i.e. re-compute the model at several yield loss levels (5%, 10%, 25%, 50%) to create a function showing cost per die as a function of yield loss. does not really make much of a sense unless you attach probabilities to those. Without probabilities simial costs would be associated with transition to .13 regardless soi or not. If you would provide me the distribution (close to impossible) then I can consider the costs re: I've seen the 4X to 6X cost numbers in several sources so did I and those numbers are not for mass-production. The only estimate of the cost for mass production came from IBM and it was substantially lower re: Now you?re just acting imbecilic... I would call it sarcastic I used ibm's estimate in cost together with the exponential function of asp. Regards -Albert

To: fingolfen who wrote (141478)	8/14/2001 4:24:26 PM
From: pgerassi	Read Replies (1) \| Respond to of 186894

Dear Fingolfen: Oxygen implantation for thin mode wafers does not cost $6000 to $10000 a wafer (that is what 4X-6X cost would be). Oxygen is cheap and the power required is cheap (no where near $10000 for one wafers worth) thus, the only cost would be for equipment. So you are implying that a production scale implanter(s) doing 30 wafers per hour would cost (30 wafers per hour * $10K a wafer * 8000 hours a year (some downtime for PM) * 1 year life (probably more than that)) $2.4 billion and that is for about 4800 WPW. I doubt the overall cost would be more than a tenth of that or $1K a wafer. The processing costs after that are little changed except for any additional defects caused by the implantation process. I find that rather difficult to see that being very significant due to oxygen is implanted at each level in the wafer to make CPUs. It is used in making the MOSFET gates, isolating components on the wafer and the area between the interconnects and between layers of interconnects. Thus the defect generation cannot be much more than 2 times the defects in any given SiO2 processing step. This makes the effects on yields upper bounded by about 2 to 3% at most. Thus the costs of a SOI CPU wafer are at most 1.5 times a normal (bulk) CPU wafer at a given process size. That is about $5 to $7.50 a good die. If it cuts power by 20 to 30% and no speed increase (far less than claimed) it would get an ASP increase equivalent to the ASP difference between a 1.2 GHZ Tbird (266) and a 1.2 GHz Palomino (MP) which is $111 (current Pricewatch). Given the above, its a no brainer to go for it. Making $100 a good die times about 200 good die a wafer equals $20K in additional gross profit per wafer. That is about $100 million a week and $1.3 billion a quarter. Using the Q2 units of 7.7 million, $770 million in additional gross profit, 60% of which falls to the bottom line (taxes) $462 million or $1.47 a share. Any company would be stupid to fail to take any step that would do that to the bottom line. Suddenly SOI does not look to be that far fetched, now does it? Pete