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.

Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?

To: Process Boy who wrote (85363)	1/6/2000 12:26:00 AM
From: Dan3	Respond to of 1572065

Re: Intel's cost per wafer, including that of labor, is much cheaper than AMD's... Allocating costs is more of an art than a science and is why accounting results aren't always reliable. Intel's operations, other than "Intel Architecture" lose a bit of money (according to, as best as I can recall, the last earnings statement, but I know I've read it somewhere recently). That's a little more than made up by profits from investments (also a recollection, I think from the last earnings statement). So I'd call the non CPU related business a wash. But, bottom line is that, the non-PC parts business neither adds nor subtracts much from the earnings equation, and if it pays for itself, its costs and revenues can be considered roughly equal. My rough guess is that about 3/4 of revenue is from Intel architecture. (28 billion revenue - 100 million units times 205 ASP is 20.5 billion from CPUs alone) If we allocate 3/4 of costs to the chip business we get costs of .75 * 21 billion or 15.75 Billion AMD has been getting about 3 billion in revenue, and we know that about half that came from CPUs (20 million units times $80 ASP for the year?) and the rest from other. Costs were closer to 3.5 billion, and most of those costs can be allocated to CPUs (3/4?) So AMD produced about 20 million chips for 2.625 Billion or $131 per chip while Intel produced 100 million chips for 15.75 Billion or 157.5 per chip. It's the best I can guess at - it's probably way off, and I don't even know in which direction! But it was an interesting exercise. Without any idea of the mix of wafer sizes at the Intel FABs, I wouldn't try to guess cost per wafer, but then the cost per chip may be at least as relevant. AMD has been paying for two FABs (really, about 1 and a half) but only producing from 1. That will change in the next few months and AMDs cost per chip should drop substantially. These are all real rough estimates, and I'd be interested in the many improvements that I'm sure can be made to them (tomorrow, I'm going to have to call it a night, now :-) Regards, Dan

To: Process Boy who wrote (85363)	1/6/2000 7:32:00 AM
From: niceguy767	Read Replies (1) \| Respond to of 1572065

Hi Process Boy: Just doesn't make sense to me that average cost per PWeeIII chip at 733Mhz or better, after factoring in all the costs associated with misfires, is less than or will be less than the average cost of an Athlon at the same Mhz level! Nor does it make sense to me that PWeeIII net revenue per chip produced, including costs of failed chips,is even close to net revenue per Athlon chip! And this disparity in net revenues is only going to widen as Athlon production levels increase thereby reducing average production costs!