*AV*--Geez, ask the tough questions?<gg>
ASYT-very little capital investment should have been made relative to 300mm programs. By this I mean, most of their products are built up from sub components and very little from in house die molds or uniques manufacturing equipment. With a little creative license, let's say ASYT is close to being the equivalent of a PC Monitor assembly line that is geared up to run 17" and 19" monitors. A good deal of the components are identical and they order the monitor screen from a sub contractor. The only real vulnerability is whether they make their own 17" or 19" external monitor cases. the real vulnerability lies with their subcontractors that may have spent money to produce a certain sub assembly.
ASYT, like PRIA, has a great deal of intelectual property (patents) on their products but for the most part, each product is nothing more than the assembly of subcomponents from a variety of external suppliers. The only real costs that should affect them are in the form of expenses and R&D for the unit designs, minienvironment designs, meeting with end customers, and software codes for the 300mm genration of products. That, in addition to any inventory of units built in anticiaption of orders. I see more carrying expenses than depreciation or underutilization of capital equipment.
the 300mm rsources of each company could very readily be redeployed to the respective 200mm programs without any effort. The real issue is the overall headcount for these companies relative to orders and revenue. there should be (not necessarily will be) a slight downsizing in the number of people for both companies given the industry slowdown. This dead time could be used by both companies to expedite existing programs and orders to customers that are waiting for programs to start or come to completion.
In a nutshell, both companies are more constrained by the humand factor rather than the computer or automation factor. Both companies are labor intensive whether it is in design, build, install, or servoce of their products. therefore headcount determines the cycle time of project completion.
Next subject. If 300mm is pushed out to 2005, as I believe, and the technology curve and fab obsolescence curves are still valid, there will be at least one new round of fab expansions prior to building the 300mm facilities. This wil be required due to technology demands, need for equipment replacements or upgrades, and capacity needs beyond the DRAM factory gluts. This will be the revitalization of the equipment industry. 8 years is way too long to expect the existing equipment set to last intact. 8 years of technology advances wil obsolete some equipment since they will no longer be capable of running the new process. And in the same timeframe, novel technologies will be developed (like Copper and CMP) that will require new types of equipment. So I see all doing well.
The primary beneficiaries have got to be the DUV players and the CVD or Etch players. I will not get into niche companies like AGAI or CFMT that may fare well but rather the companies that will make up the industry standards.
On first glance, CYMI, ASMLF, DPMI, PLAB, NVLS, AMAT, KLAC seem to be the best pure plays on the front end and TER on the back end. This list could be expanded to include ASYT, PRIA, ATMI, IPEC, SFAM, CFMT, LRCX, KLIC, VECO, UTEK but why venture here when the first 7 should lead the pack.
If forced to pare down the list, I would just have CYMI, DPMI, NVLS, AMAT and KLAC as a mini mutual fund. As a dark horse for group 1, I would be remiss ifI didn't state that as much as DUV litho will lead the way, the other members of the group 1 list will bring on technologies (copper interconnect) that will rely very heavily on the hardware and chemicals supplied by ATMI. ATMI may get dragged along with the introduction of new device technolgoies requiring their unique product offerings. ATMI has hardware relationships with a few of the group 1 companies.
The copper process will move forward on the 200mm platform. as a matter of fact, it might be easier to implement on 200mm since machine induced uniformity issues are probably minimized on the smaller diamter wafers.
If the announcement is made properly, it should help a good many of the companies. The entir financial community is intimately aware of the financial burdens that the 300mm programs were placing on all companies. The re-engineering effort is astronomical and I am sure was factored into the prices of many of these stocks going forward. With the burden of these programs off their books and with a resurgence of 200mm orders, they should be able to present a rosy picture. Remember, we have seen gloom and doom in the past year where the words "200mm programs are falling off more rapidly than expected in light of the expected ramp of the 300mm programs. Why build a 200mm plant when, if you wait, you could go 300mm and be much more cost effective". Well, if you state 300mm is dead for 8 years, you force the hands of those sitting on the fence. The 4",5" and 6" fabs that were stretching the envelope to make the big jump all at once, must now go rethink their strategy. The announcement, if and when it comes (I believe it is forth coming) will force companies to move forward on their stalled 200mm fabs. It might even force them to build bigger fabs to allow for capacity expansions that would allow for the equivalent amount of wafers to match the output of a 300mm fab. (equivalent ratio of
150mm * 150mm)/(100mm * 100mm or 2.25 more area)
This 2.25 ratio of increased area is extremely significant when you make the preceding comparisons of 6" to 8" which is a ratio of only 1.78, which was significant.
there will probably be some negative fallout since the market is inefficient. All 300mm work was written off as expenses or R&D already so it is water under the table.
Andrew |
