SI
SI
discoversearch

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 Applied Materials
AMAT 230.77+0.9%Nov 12 3:59 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: robert b furman who wrote (50491)8/14/2001 12:37:45 PM
From: mitch-c  Read Replies (1) of 70976
 
I've been thinking of it as a commodity oversupply problem, as the DRAM makers are seeing now. When supply expands faster than demand does, prices drop quickly. In effect, a company gets punished for executing well.

We're seeing the same thing with CPU's - the processing power of CPU's has developed much faster than Moore's Law predicts in recent years, primarily because of higher clock speeds. Software development hasn't kept pace. When Intel took a huge cut from the price of their high-end CPU's, I saw that they're trying to dump rapidly depreciating inventory - a commodity-style supply/demand imbalance.

In the SCE market, the die shrinks and wafer area growth are both happening at the same time. The chips/wafer ratio (or machine throughput) is growing comparatively faster than demand (especially in this economy). You can get 3x to 4x the number of chips per process cycle out of a newer machine. (Granted, that can reduce your overhead a bit through economies of scale, but still ...)

Combine all of the above, and I just don't see a *current* market for capital equipment that allows greater throughput, when the profitability of the chipmakers is already impacted by overproduction and excess capacity problems.

Now, some disclaimers. This is a static analysis of a notoriously dynamic (cyclical) industry. This is where the target sits today; now we have to figure out where it's moving, so we can figure how to lead it.

First, the lead time for SCE orders is a high fraction of recent cycles. Chip companies have to anticipate their future needs at a time when they are least able to pay for them. (i. e. Now.) Prediction in a chaotic environment is almost hope and guesswork. They compensate by multiple bookings and/or postponing deliveries, so Morgan's "all at once" comment is a solid, shrewd analysis of his customers' behavior.

Second, this presumes a fairly predictable smooth (linear or exponential) demand for chips and a highly variable supply. A discontinuous event could knock the current supply/demand relationship out of kilter quickly. (Possible discontinuities - WinXP? Wireless everything? Japanese earthquake? Bay Area earthquake? China invades Taiwan?)

I hope this gives you a look from my perspective - right now, it feels like the entire chip-and-related industry is pushing on a supply string when the change really needs to come on the demand end.

- Mitch
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext