The Right Questions

The Right Questions "Equipment Life Cycle Management" will be addressed in a panel discussion at the Strategic Business Conference this week. This is a subject that is near and dear to my heart. The content is quite relevant to the post I made Thursday (the post is below this message - it's called "Capacity Observations"). Fortunes have been made procuring and selling the excess equipment produced during the bubble years. It's not just the excess production that generated this wealth, equipment from fabs that were shutdown was also purchased by distributors and channeled to many different companies in the chip manufacturing industry. Most in the industry know that China's device makers have absorbed a lot of these tools. One could even go as far to say this market is really the driver of the semiconductor manufacturing ramp in China. The quantity and even the capability of the tools moved through these channels should not be under-estimated. Complete fabs have been auctioned and then resold. As one example, Motorola shut down more than 20 facilities, both back and front end, during their reorganization. There are many, many others that have done the same. OEM capital equipment companies have also been moving tools into the market. Some of these are older tools but a great percentage are tools that were used in development labs. For many device manufacturers development lab tools are a great deal because their capabilities often exceed the process requirements used in their fabs. Personally, I think device production from this "recycled" capacity is greatly under-estimated by many that follow the industry. I also believe that this business is something that has a great influence on semiconductor capital equipment spending ratios. Why? Because the equipment sales numbers used to calculate capital spending ratios is based on sales reported by global OEM tool manufacturers. The numbers do not include sales made by entities that buy and resell equipment purchased at fab auctions. The numbers also do not include the tools that refurbishment houses send back in to the market. Recent announcements from OEM capital equipment companies make it pretty clear that this older, but still production worthy market, is something that must be acknowledged. That interest is what prompted the SBC Committee to put the "Equipment Life Cycle Management" panel discussion on the agenda. Here's a backgrounder on the discussion that will take place at SBC: Today, approximately 40% of global device production capacity is on 6 inch or smaller wafers and >20 year old device technology (>0.5u). And, 64% of all fabs globally are >8 years old, with global average fab life of 12.8 years. These facts present some difficult challenges to the Semiconductor device manufacturers and the OEMs that supply the equipment for the fabs. Did we really expect that these conditions would prevail 20 years ago when those earlier technologies and equipment designs were first installed? Were the Semiconductor manufacturers' asset depreciation models based upon fab and equipment lifetimes of 15 to 25 years? Were the OEMs' support strategies designed with this extended life-cyle in mind? What is the true market life of a device technology? What is the true useful life of semiconductor capital equipment? Clearly, in the case of the newer leading edge fabs the focus is on CapEx and ROI. In the case of the more mature fabs, the focus moves to OpEx. Faced with these realities, what are the strategies that the device manufacturers and the OEMs should pursue to insure the greatest return for their customers and their shareholders? Great questions.... Let's toss in one more, contributed by one of the panel participants representing a device manufacturer: What other suppliers come into play with mature equipment in addition to the equipment OEM and the original leading edge device manufacturer? These could include other device manufacturers with different product portfolios, third party parts and service suppliers, licensed (or unlicensed) third party equipment suppliers, etc. There is a whole different value chain here other than the original one. Lots of things to consider here. I know, I use this quote from Bob Dylan a lot but it seems very appropriate, "The times, they are a changin'" In the publications I send to readers (subscribers) and the various blogs at this site I am going to write a lot more about this. It's a huge issue.

---

Silicon Shipments: SEMI Q1 Data Another item related to my "Capacity Observations" post last Thursday hit the wires late in the week: SAN JOSE, Calif., May 5, 2005 – Worldwide silicon wafer area shipments decreased less than 2 percent during the first quarter 2005 when compared to the fourth quarter 2004 areas shipments according to the SEMI Silicon Manufacturers Group (SMG) in its quarterly analysis of the silicon wafer industry. Total silicon wafer area shipments were 1,465 million square inches during the most recent quarter, down from the 1,486 million square inches shipped during the previous quarter. The new quarterly total area shipments are 4 percent below first quarter 2004 shipments. "Coming off of a record year, we expected and experienced a slight drop in unit shipments during the first quarter of 2005" said Makoto Tsukada, chairman SEMI SMG and general manager of Shin-Etsu Handotai Co., Ltd. "However, we continue to see strong sequential growth in 300 mm wafer shipments."

Quarterly Silicon Area Shipment Trends Silicon Shipments- Millions of Square Inches

	Q1 2004	Q4 2004	Q1 2005
Polished	1,130	1,114	1,107
Epitaxial	336	314	307
Nonpolished	63	58	51
TOTAL	1,529	1,486	1,465

The numbers for next quarter, I suspect, will be even more interesting.

---

Comments and questions are welcome.