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 : ASML Holding NV -- Ignore unavailable to you. Want to Upgrade?

To: etchmeister who wrote (1176)	7/27/2006 11:39:41 AM
From: niek	Read Replies (1) \| Respond to of 42736

TSMC: Foundry slowdown? What slowdown? Thursday, 27 July 2006 FabTech In various quarterly conference calls that are currently ongoing some financial analysts have been questioning executives over a slowdown in capacity ramps at major foundries and a subsequent drop in capital spending that could occur in the second half of the year. Analysts fears or pessimistic opinions which ever suits you, have been fuelled by the softening of end demand in 2Q, inventory build and belief that demand may not spring back strongly enough in the second half to kick the dust of potential oversupply. Key parameters are fab capacity ramps and capital expenditure plans for the next two quarters. However many analysts have noted that wafer prices in a tight capacity environment seen all year have not resulted in ASP's edging up, rather they have been edging slightly down. Some have said that this indicates oversupply looming, reinforcing pessimistic views that a peak has passed and it's down hill from here! The problem I have with this view is that it doesn't take into consideration current foundry business dynamics or general IC unit ASP dynamics. With the old duopoly (TSMC, UMC) foundry business now long gone, replaced by the gang of four (TSMC, UMC, SMIC & Chartered) competition at more advanced nodes (130nm and below) has become more intense. All foundries are finding it harder to get premium prices for leading-edge production compared to 5-years ago. The issue with IC unit ASP's is that they have been trending down for over a year and according to Malcolm Penn at Future Horizons, a dedicated semiconductor market research firm, ASP's take a very long time to change direction even when the dynamics show that ASP's should be moving up, they don't respond at will. So, falling foundry wafer prices are as much to do with competition as weak demand or over-supply. In fact this is a benefit to the industry as foundry customers have more choice where to fabricate chips and in a period of tight foundry capacity can shuffle production via agreements with two or more foundries, preventing double ordering and a definite over-supply scenario. Indeed we now have the likes of ATI, Xilinx and Qualcomm being able to use a number of foundries for production using leading-edge processes, something that was very hard to do without using TSMC as the main supplier. So, putting wafer ASP's to one side are there problems looming in fab capacity slowdown and capital expenditures? According to the data provided by TSMC that takes into account the rest of this years plan, it seems that it is business as usual. Capacity ramps are continuing at both 200mm and 300mm fabs (see graph below), while CapEx is remaining at the April updated level of between $2.6-$2.8 billion US dollars. Year to date TSMC has spent $997 million, which shows that demand is expected to continue to expand in the second half and into 2007 with such a conservative spend so far. We could look at this another way (glass half empty) that indicates that should foundry demand not meet TSMC's expectations in the second half it has the opportunity to cut CapEx easily as more than half the spend is second half year loaded! However this doesn't hold much weight in my opinion as it fails to take into consideration that at the macro level semiconductor demand is still expected to show good growth for the year and its second-half loaded as usual. Not only that but 2007 is still expected to show good growth as world GDP projections remain strong for all regions, something that doesn't happen all the time. Fundamentals here really shouldn't be ignored. TSMC may want to project some cautionary tales as visibility for a foundry really isn't that good, but it is also playing safe so that financial performance expectations can be handled better. Therefore it is saying one thing but capacity ramps and CapEx spending continues, indicating strongly something else! (Chart Analysis) Fab12 and Fab14 are 300mm fabs. Fab12 is nearing full capacity with its ramp rate at or near the 300mm fab average ramp rate figure of between 2,000wspm and 2,500wspm. Fab 14 has been in a continuous ramp since 2005 and through 2006, according to TSMC's figures. The ramp is far more aggressive than Fab12 this year. Indeed Fab14 is ramping between 6,000wspm and 7,000wspm in the next two quarters, right in the middle of the best fab ramps so far achieved on a monthly level for logic based fabs. The other fabs are TSMC's main 200mm facilities that have been earmarked for extra capacity ramps in 2006. Both Fab 6 and Fab 8 will be at new full capacity levels by end of 2006. The Shanghai, China facility has seen sporadic but progressive ramp in 2006 and should reach 30,000wspm by year-end. The ramp rates of these fabs suggest that TSMC is continuing to see demand continue to increase through the rest of the year both for leading and trailing edge devices across many markets. The chart: fabtech.org

To: etchmeister who wrote (1176)	7/27/2006 2:42:58 PM
From: niek	Read Replies (1) \| Respond to of 42736

re: flaky article This guy is really a non-item. Non-Linear Progression By Stephen D. Simpson, CFA July 26, 2006 Given what I wrote earlier today about ASML, and how the semiconductor industry seems to be getting harder and harder to forecast, guidance from Linear's management was rather comforting in one sense. It said that the outlook for the next quarter was pretty murky. Not necessarily "women and children to the lifeboats" murky -- more like "don't expect a blowout" murky. At least I'm not the only one whose crystal ball has turned into a cueball.