To: lrrp who wrote (1047 ) 9/9/1998 12:01:00 PM From: Ian@SI Respond to of 2313
Lrrp, It would have been much easier if you would have just asked me the meaning of life, the universe and everything. Then I could have quoted Doug Adams and told you, "The answer is 42". and then let the thread help you out when you ask, "But what is the question?". <g> I have never smifd a fab ( sounds like what my dog does in the backyard); Actually, by your definition, the Street SMIF'd ASYT. Sorry, I coined a verb which saved a 2 sentence description. When I talk about SMIF'ing, I refer to making all wafer movement processes between process tools happen in a SMIF pod.I am wondering what the problem is Chipmakers are facing a larger deflation rate than they've seen in the last decade and possibly ever. The traditional deflation rate of 30% a year for their products would look like heaven to each and every chipmaker. <What's the Peter Lynch story?> Impurities cause fab yield to fall. Removing impurities from the entire clean room (many 100,000s cubic feet) is expensive and insufficient and becomes moreso as feature sizes reduce. Even the smallest impurities can cause yields to fall. Asyst's minienvironments provide a "mini clean room" for each wafer or cassette of wafers. This provides a lower cost, more effective means of improving 'clean room' effectiveness and raising yield. In addition the SMIF pods have hardware (SmartTraveller system) and software which is used to prevent misprocessing of wafers. "SMIF"ing an existing fab can extend its useful life by about 6 years at a cost of about $15M. i.e. - A savings of about $1B to $1.5B compared to building a new fab with similar clean room characteristics. I can't tell a pod from a minienvironment neither. A SMIF pod is a minienvironment that holds a tray of wafers. Jenoptiks wants to eat ASYT's lunch. So far they've been on a starvation diet. May they starve to death. No offense meant, nor any ill will. ;-) PRIA doesn't really compete with ASYT. ASYT tried to compete with PRIA and quickly repented for its sin. It no longer tries. But Mihir has a company with which he might try once again. ... and he may very well get his butt kicked once again. PRIA takes care of the long haul transfer of wafers between clusters and does some other stuff, most of which ASYT doesn't do. Other Competition that you don't know about? Probably. But I don't know about most of them neither. How many angels can dance on the head of a pin? You should ask Alastair Glass, head of photonics research at Bell Labs. Using MEMS technology, his folk have put 100 micro-mirrors on the head of a pin which can be controlled to add or drop specific wavelengths to a DWDM optical fiber. i.e. - a prerequisite to developing an all optical switch that essentially takes zero time and zero processing power to switch Gigabits/second of bandwidth. He could probably give you a much more informed answer than I. Good luck, Ian.