Larry..Thanks so much! Let none accuse you of failing to explain :-)
And all worth reading. Your DSP experience certainly has much useful applicability for us, though we might be more interested in ASICs and DRAM applications... BTW... Several points we might clarify, but first let me drop a current DPMI chart on the scene here:
207.95.154.130
Regarding the chart, we could use a nice stochastics crossover to start a run with a potential to about 54/56, IMHO. But DPMI is not yet moving with some of the other semi-related and or/tech stocks...ALTR, ADI, IDTI, IMG, ASMLF, ZITL, XLNX, for example...perhaps "recouping" from that bubble pop which came with the UT announcement :-)
Wouldn't you be quite surprised (as I would be) if the industry were to be using the new UT/Sematech/ISI/CYMI/DPMI process for extending DUV below .08u anytime in, say, the next 6 or 7 years? If used at all, might they use the new process for a single pass with a specialized mask to define specific features while the rest of the wafer is processed with more conventional DUV processes?.. But what impressed me about DPMI's involvement in the UT experiments is that it places DPMI in the (apparent) technological lead, just when photomasks will increasingly be done by large specialist companies, and less and less by inhouse shops or "mom and pop" shops.
Second, the significance of extending the use of photomasks to/beyond .08u is that it extends the life of the Photomask companies at least out another 15 years...just when their existence was being called into question due to progress in x-ray, electron beam or other "direct write" lithography processes. Those expensive processes now must await some new cost breakthrough before conventional optical lithography methodologies are threatenened, IMHO
Third, as we have been discussing on the DPMI thread, the complexity of chips (and the complexity of masks) goes up, not linearly as feature size is reduced, but exponentially. This week we finally got the pricing for photomasks, and it looks even better than I had understood. While current masks run about $4700 each, the cost jumps to $21,000 each as feature size goes to .18u, and jumps again to $47,000 when the masks must deal with .13u. (One must use some imagination to consider .1u or .08u. data not avail). The key point here is that DPMI can experience very sharp revenue gains on little unit volume increases for chips, depending more on the rate of advance in technology than on the rate of increase in unit volumes.
Now...How will DPMI come through the bottom of this semi cycle...if they are able to come through with minimal damage in the current quarter, it would be a major boost for those who do their investing by looking through the front windshield... Will be interesting. More to come, we can bet!... Appreciate your inputs, Larry. Again, welcome, and hope you'll continue sharing your thoughts investment decisions get made over the coming weeks...
"The Saintly One" --- TSO |
