|
SEMATECH softens tone on 450mm wafer transition |
|
|
|
May 22, 2006 at 04:18 PM |
 Getting industry consensus on the next wafer size being 450mm is becoming more crucial than that required for the move to 300mm wafers. Currently led by Intel Corp through its membership of the International SEMATECH Manufacturing Initiative (ISMI) the organization is having to prepare extensive quantifiable data that will need to be used to persuade not only other chip manufacturers of the viability and necessity of another wafer migration in 2012 but bring inline wafer growers, equipment vendors and trade associations regarding automation standards and such like.
According to ISMI, its members have reached consensus on key concepts of a productivity improvement roadmap for moving the semiconductor industry toward 450mm wafers. Without the improvements in productivity using 300mm wafers it is expected to be a lot harder to convince enough chip manufacturers to commit to the larger wafer size.
Therefore ISMI will undertake extensive work on assisting 300mm chip manufacturers to gain significant improvements in productivity over a yet to be specified period, which come under the name of 300 mm Prime (300P).
"The Guiding Principles will be used as a foundation for gathering specifics on productivity designs, attributes and metrics to be pursued in today's 300 mm fabs, which will help pave the way to 450mm implementations," said Scott Kramer, ISMI director.
However, equipment vendors have yet to be convinced of the commercial viability of developing tools for the larger wafer size. Mike Splinter, President and CEO of Applied Materials and the largest equipment supplier to the semiconductor industry had deflected questions from financial analysts on this subject. The financial community is concerned about the significant RD&E costs that it took to develop 300mm toolsets and 450mm would seriously impact the profitable operating margins of the company.
Other equipment executives have also raised doubts about the migration though some see the inevitable change significantly past the date outlined in the ITRS of 2012.
The softening in tone was seen in the press release comments of Joe Draina, ISMI associate director: "300 Prime is a strategy for improving 300mm productivity while bridging to eventual 450mm manufacturing. This strategy draws on lessons learned from our industry's last conversion from 200mm and recognizes the importance of collaboration, consensus-building, and compromise among chip-makers and equipment suppliers."
ISMI is now attempting to demonstrate to the equipment vendors that things are very much open to discussion rather than moving in a direction few chip manufacturers other than Intel have requested.
The study work for the transition now includes references to the design and configuration of tools and fab automation infrastructure being ‘forward-compatible' from 300mm to 450mm.
Though it is yet to be known whether scaling of this type can be achieved ISMI had made it clear that it will have to be proven and tested in a real world pilot scheme to mitigate the concerns and risks very much at the fore for equipment vendors.
ISMI has also conceded that detailed and comprehensive business modeling of using the larger wafer size for a typical range of chip manufacturers products, markets and manufacturing needs will need to undertaken.
Significantly it would seem ISMI is ready to attempt to alleviate concerns amongst major foundries, for example that deal most with "high- and low-volume product mixes; and multiple flexible lot sizes operating within acceptable cycle times," as stated in the press release.
A key aspect of any migration will be the support of the major foundries as IDM's continue to produce both in-house and outsource production and there is a need for comparable costs. Furthermore the foundries are seen as major purchasers of leading edge equipment and without their agreement, equipment vendors would be more reluctant to build the new tools as the ROI available market would be reduced significantly.
"The introduction of 300P is designed to be an iterative and self-testing process that will generate immediate, mid-term and long-term benefits for the industry," Kramer said. "We see it as a gradual and affordable transition that is responsive to its participants and adaptable to their needs. As I've said before, this process is a marathon, not a sprint."
|
News 
