Semiconductor Industry Association Roadmap: Blueprint for tomorrow

12/01/1997

Semiconductor Industry Association Roadmap: Blueprint for tomorrow

Paul Nesdore, Solid State Technology

Solid State Technology spoke with Owen P. Williams, chairman of SEMATECH`s Roadmap Coordinating Group, about the Semiconductor Manufacturing Roadmap published last month by the Semiconductor Industry Association.

SST: Everyone was looking forward to the new Roadmap. Why is this such an important document for the industry?

Williams: The Roadmap represents a consensus for the entire semiconductor community. This includes not just the industry itself, but the university and government representation. Part of our plan is to get the broadest, widest participation as possible. This past May, we invited 650 participants from all walks of life to come to Austin to critique the plan. Of those invitees, 330 showed up.

SST: Wouldn`t the industry do pretty much what it needs to do anyway, that is, meet competition, without the need of a Roadmap? Why do they need it?

Williams: There are two basic reasons for the Roadmap. One is that it gives the researchers a 15-year look at what requirements their technology is going to have to provide. So, if you are working on a technology that will satisfy the needs 10 years out, you will have a good indication of what kind of requirements that technology will have to perform to. This has never been provided before. Prior to the Roadmap,the researchers were guessing, and often times didn`t know that the technology they were working on would not meet the requirements.

The second reason is a short-term benefit in that it gives an industry consensus of what the short-term manufacturing tools will have to perform to. So they have very good direction in producing their equipment along with the high confidence that that`s what the industry will need. So the equipment suppliers and the researchers both benefit.

SST: The Semiconductor Manufacturing Roadmap is unique in that many other technology areas, like computer software and hardware, do not produce this type of detailed industry roadmap for their development efforts.

Williams: Yes, it is unique. This approach started with the Semiconductor Research Corp. almost 15 years ago. For a long time it existed by itself. Then about 10 years ago, we generated SEMATECH, which is also an industry consortium. After that we had a lot of activity in the US government looking at where there were additional needs necessary to preserve the semiconductor industry in the United States as a reaction to government-funded programs in other countries. At that time, a national semiconductor advisory committee to the president was formed. This report generated a challenge to the government to fund a sort of a leap-frog development program. The problem that these three events generated was that there were three different people going to the US government saying, "You need to do my program to save the semiconductor industry."

We realized that we needed to speak with one voice and that voice had to be industry-wide consensus. In order to do this, we started to generate the Roadmaps. We put the first one out in 1992. We renewed it in 1994, and now again in 1997.

SST: Do you feel the Roadmap needs to be updated more frequently, less frequently, or about the same?

Williams: We think we`ll go back to every two years now. We went three years this time, and it was too long. The technology moved too fast in that period.

SST: Are there negatives in producing the Roadmap?

Williams: One negative is the possibility of a wrong interpretation. This happens when people just glance through the Roadmap, or hear about the Roadmap and don`t really read it, sometimes interpreting it as having all the solutions. So they ask why do we really need to fund any new research. The Roadmap is a roadmap of needs and requirements, not a roadmap of solutions. So the downside is a misinterpretation of it.

The second downside is that the funding agencies that pay for research perceive that the only known solutions are those that should be worked on. We strive in every case to try to indicate that we are putting these challenges out there to generate creativity, to find new ways to solve these problems. But, often times, it`s perceived that these are the only ways to do it, so funding agencies won`t fund any out-of-the-box thinking.

SST: Like in nonoptical lithography?

Williams: Sure. Our Roadmap has nonoptical in there, and we do list potential solutions, but not all the potential solutions. And if you read the text, it tells you that these are there only to let other researchers know what is being addressed in that arena. Too frequently, the funders interpret this as the only thing to sponsor, and that`s a downside.

SST: Has the Roadmap evolved over the years in what it specifies or the details that it specifies?

Williams: Yes, it`s getting more specific and it`s getting more concise in its presentation.

SST: Will this trend continue?

Williams: Yes, I think we`ve done a much better job this time in correlating all the consistency of the requirements.

SST: Is there something that the Roadmap should become that it isn`t now?

Williams: I think it`s complete. The other area that should be addressed is how to fund research, but this is outside the Roadmap. That is, how to use the Roadmap to make sure that we fund and implement the research that we need to meet the requirements. But the Roadmap - as it is - is pretty good. We have specifically kept the Roadmap working groups away from getting into the politics of funding and such things. We wanted the Roadmap to represent pure technical input. So we have steered away from any suggestion of what program should have priority and topics of that nature. The Semiconductor Industry Association has an implementation working group that is trying to strategize on how to implement the Roadmap. I believe that needs expanding this time around, but not as part of the Roadmap.

SST: Is there a last thought you`d like to leave on?

Williams: Yes. There is one point I would like to emphasize. That`s the problem of misinterpretation. The Roadmap is not intended to stymie creativity or to pick winners or losers.

SST: Thank you for your time, and we look forward to reading the new Roadmap.

Owen P. Williams can be reached at Motorola, 6501 William Cannon Dr. West, Austin, TX 78735-8598; ph 512/891-3749, fax 512/891-3785.