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 : LSI Corporation -- Ignore unavailable to you. Want to Upgrade?

To: shane forbes who wrote (15201)	9/22/1998 3:48:00 PM
From: patrick tang	Read Replies (1) \| Respond to of 25814

I think investment sentiments are beginning to change: between yesterday and today, tier 1 TXN, STM down 5% to10% tier 2 XLNX, ALTR, ATML, LSI, ADI, VLSI up 10% I am almost willing to bet money is starting to flow from tier 1 to tier 2s. Translated, are investors are starting to look past the next 2 quarters? patrick

To: shane forbes who wrote (15201)	9/22/1998 8:59:00 PM
From: Grand Poobah	Respond to of 25814

In many of the fabless companies' 10-Qs I see the argument time and time again about product development costs to get qualified at foundries; also the time/costs involved in moving to new fab processes or design shrinks. These costs should not vary that much between fabless and fabbed companies. The big difference is in wafer costs, which vary with supply and demand. These can especially be a differentiator at the leading edge geometries, where supply in the foundry market is often constrained (although TSMC is pushing to change that). Product qualification is pretty standard whether it is done for a foundry's process or your own internal fab process. It is driven by the customer and is done according standards from organizations like JEDEC. These are the same for everybody across the board with minor variations. The problem with moving between foundries was a lot more serious several years ago when the foundry industry was much less mature. The foundries were small and could not support larger customers, but that has changed to a large degree. It is still a factor, but to a lesser degree. Also, design shrinks done at a foundry are pretty similar to those done in your own fab. In your own fab you do have more control over the process, but my experience has been that the foundries (at least the better ones--TSMC, for example) do a better job of controlling their processes than a lot of captive fabs do. Maybe the reason is that the foundry's paycheck is more directly dependent on their customer's happiness than an internal fab. Of course I'm sure there are exceptions, but that's my general observation. I'm not saying that having a captive fab is a bad idea. In fact, for a company the size of LSI, it is pretty much a necessity. However, I don't agree that a fab owner necessarily has a big advantage over a foundry customer. Sometimes it is a disadvantage (when there's a glut of capacity) and sometimes it is an advantage (when wafer supply is constrained). Over the long haul it probably averages out to a small advantage cost-wise. However, the stability of supply provided by your own fab is a necessity for the biggest companies and probably the biggest reason to own a fab. Regards, G.P.

To: shane forbes who wrote (15201)	9/23/1998 6:57:00 PM
From: Richard H.	Respond to of 25814

From DLJ Semiconductor Equipment Book-To-Bill Last night the August data for the three month rolling average book to bill ratio was released. We have hit another record low, the overall book to bill ratio was 0.60 compared to last months 0.69. The front end book-to-bill ratio (AMAT, NVLS, LRCX etc) was 0.61 and the ratio for test/assembly (TER, CMOS, BESIF) was 0.55. There is a slight slowing in the downward trend as August was down only 15% compared to July levels which trended down 20% from the sequential June level. Overall the numbers continue to reflect the industry's downward momentum with some slight deceleration in assembly/test. Though we have seen a firming in DRAM pricing and some spot processor shortages, they are not signs of an impending recovery. We are still grossly oversupplied in DRAM and have more than enough processor capacity coming on line. We cannot have a true recovery until we start to get closer to full utilization of existing fab capacity. We remain on the sidelines while we still have all this negative momentum and don't even want to consider changing that opinion until we at least start to see some sort of a bottom.