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: Petz who wrote (91223)	11/27/2002 7:55:29 AM
From: Road Walker	Respond to of 275872

Message 18279167

To: Petz who wrote (91223)	11/27/2002 9:04:57 AM
From: Dan3	Read Replies (3) \| Respond to of 275872

Re: Intel losing $200M/quarter selling flash It's an interesting exercise to calculate costs in a capital intensive sector that has seen a collapse in sales. I'd collected this FAB list earlier: Message 16829926 from various sources. IIRC, Intel converted (or started to convert) one large flash FABs to .13. and built another from scratch. Only Intel knows the costs for that conversion, but $1 Billion and $2 Billion is probably about right. In the same period they maintained other flash FABs, but bottom line is that on a 5 year depreciation schedule capital costs for those two FABs alone works out to about $150 million per quarter. Capitol costs from the other FABs are probably half that, say $80 million, per quarter. (FASL, which has been much more conservative in FAB construction, has capital costs that are about 1/3 to 1/2 Intel's). If operating 4 Flash FABs, and testing and packaging their output, is costing Intel $80 million, per FAB, per quarter, and and their flash R&D and marketing efforts another $20 million, then the total could be as high as $570 million. But that loss estimate could also have reflected a lower revenue base for Intel, so that loss could have come from a lower revenue base. Intel may have lost a number of flash contracts to AMD for the coming quarters. There are a number of markets that Intel is finding itself locked out of because of Stratflash's unreliable technology. AMD's mirror bit Flash works on the basis that a cell in a microchip can be charged today, and then that read 20 years later to see if it was charged or not - it's digital. Intel's stratflash works on the basis that a cell can be charged to 1 of 4 levels, then read 20 years later to see which level it was charged to - it's an analog/digital hybrid. But loss rates vary depending on various conditions, including temperature. When GM, or Honda, etc. program the engine control computer on one of their cars, they don't know if it will spend the next 20 years on the north slope of Alaska or the desert in Palm Springs. Average temperature of 0 or 100. To a lesser degree, the same holds true for cell phone manufacturers, etc. So AMD's flash technology is a much more reliable choice, and this is why AMD mirror-bit flash is guaranteed to hold its data over a wide range of temperatures for 20 years, while Intel can only say that strataflash can hold its data for "up to" 20 years. Things are looking pretty good for AMD's flash business!