Spring 1999 Newsletter

Beyond the Roadmap

Robert Dutton

Photo of Robert Dutton I have received a significant amount of comment and feedback in response to my article entitled "Off the Roadmap Research in Semiconductors" in the Summer 1998 Newsletter. In fact, several remarks were sufficiently compelling that an ongoing dialog seems in order. The bulk of the responses have come from, and are centered around, industrial perspectives. Hence, I would like to use this input to focus the following comments.

Before discussing specific observations, I'd like to share with you the venue of a recent workshop, jointly sponsored by NSF and SRC, on the topic of "Nanotransistors: Technology, Physics and Simulation"
(www-tcad.stanford.edu/tcad/descartes/descartes.html). The objective of the meeting was to identify critical long-term issues, directions, and strategies for nanoscale transistor research, and approaches to maximize the impact of computational electronics on advanced device research. The meeting focused primarily on four topics: (1) What are the important problems? (2) What is the underlying theory? (3) What computer tools are needed to meet the challenges? (4) How should we (industry/academia and computationalists/experimentalists) collaborate and share results?

Many issues about what's beyond the end of the (NTRS) "roadmap," and especially challenges that the university representatives see as worthy research topics, were discussed. The meeting sparked a lively discussion and generated many thought-provoking questions. I encourage those of you with interest to visit the summary, posted on the Web.

During the workshop, Bob Doering (Texas Instruments) expressed Industry's concern that the roadmap may be getting in the way of attracting university talent to the semiconductor field and that the notion: "Roadmap, ergo est" needs to be dispelled. In a similar vein, Ralph Cavin III (VP, SRC) wrote me an extensive letter seeking to both understand the manpower (students) and faculty "mind share" issues, and to articulate challenges and contributing factors (industry progress in scaling and federal funding shifts) that SRC specifically faces in funding. I was especially excited (and pleased) with the following closing comments from Ralph:

...the semiconductor industry is entering domains where traditional assumptions no longer apply and where extraordinary innovation is needed. In a very real sense, the industry will need to reinvent itself over the next decade. We feel that the opportunity to invent at the granularity of matter is a challenge worthy of our best minds and we seek your support in our quest.

Coming back to the joint NSF/SRC workshop, Jim Plummer, speaking from the perspective of experimentalists, commented on student interests in long-term, curiosity-driven problems. Moreover, the fact that students want problems that they can view as a complete (and potentially system-oriented) problem, as opposed to a "piece of a problem," puts a premium on attracting the best and the brightest. From this perspective, the emerging areas (i.e. bio-/human-interfaces, MEMS, nano-tubes, etc.) allow universities to build new applications for chips and to define new (follow-on) industries that may shape a very different future for humanity, including the engineering talent that serves those needs.

Given the reality that these "off the roadmap" areas are where the action is from the perspective of government funding agencies, there is an ongoing need for the semiconductor industry to consider how to create synergy with this paradigm shift.

The challenge of funding for future semiconductor research, especially the experimental means to explore options, is the key question. We are pleased to announce once again the launch of two MARCO (Micro-electronics Advanced Research Corporation) Focus Research Centers: Interconnect Technology and Design and Test. Based on several years of CIS seed funding for key Stanford faculty, across a spectrum of topics (i.e. systems, circuits, design, technology, etc.), we have garnered strong participation in both of these programs.

"In a very real sense, the industry will need to reinvent itself over the next decade."

-Ralph Cavin
SRC Vice President

One of the more exciting aspects of the new FRCs is the network of participants (both inter- and intra-university connections) that to some degree mimics the DARPA-style synergy created by multiple connected contracts. In considering strategies to launch "blue sky" research in possible breakthrough technologies, there are two recent data points from the Stanford environment that warrant brief review: Stanford Graduate Fellowships (SGFs) and the NSF-sponsored National Nanofabrication Users Facility (NNUF). Stanford's commitment to fund the best and brightest, without a preset research agenda, is showing exciting progress. The topics these SGFs are selecting, and the very positive and productive support/mentorship coming from many of our CIS partners who have joined in funding them, position the concepts of FMA for a much longer-term perspective. Similarly, the innovative and highly diverse use of the NNUN generally, and Stanford's Nanofabrication Facility (SNF) specifically, reflect nano-diversity on an impressive scale.

As a final note, this article has highlighted an alphabet soup of research activities that are changing the face of the semiconductor research agenda -- SRC, SIA/MARCO, NSF (DesCArtES and NNUN), and SGF. While generally there are many signs of "good news," in terms of progress defining this agenda, we as a community -- both academic and industrial -- need to continue actively promoting a higher level of awareness and urgency rebuilding support for this critical economic sector.