Is It Time to Dump the Roadmap?

What's not in this issue, but is on our web site, is an interesting opinion piece by Dr. Henry Smith, Keithley Professor of Electrical Engineering at the Massachusetts Institute of Technology (Cambridge). He has got me thinking about the International Technology Roadmap for Semiconductors (ITRS). I don't want to steal his thunder, and I urge you to read the whole piece, but what Dr. Smith discovered in a recent trip to Japan is that there is still significant merit in a technology that has now been all but abandoned outside of Japan: proximity X-ray lithography (PXL).

Dr. Smith will be presenting more details on why he believes this technology still has life in our February issue. But he says that, at the XEL 2000 Conference in Yokohama (Nov. 13-14), where progress on all four next-generation lithography (NGL) technologies was reported, it was clear that Japan's PXL is many years ahead of the other NGL technologies, and would have a lower cost-of-ownership.

"As a result of over 30 years of development, all of the issues, and any potential problems, with PXL are well known and solvable," Dr. Smith says. "In contrast, the problems of EUV and EPL are only slowly being revealed. It is not at all clear that either could succeed in semiconductor manufacturing, regardless of how much money and manpower is devoted to them. It now appears that PXL could be introduced into manufacturing on a short time scale and persist as a cost-effective technology all the way to the end of the silicon-semiconductor era, at about 25 nm channel lengths."

I don't know if I agree with Dr. Smith - putting a synchrotron ring in one's backyard never seemed an elegant solution to me - but it raises an interesting question. What if, despite all their wisdom, those folks who developed the ITRS were wrong when they decided to focus on 157 nm optical projection, extreme-ultraviolet (EUV) projection and electron-projection lithography (EPL) ... and not on PXL?

The advantage of having a roadmap is obvious: It gets everyone focused on the key issues that need to be addressed and puts limited resources to work in the most effective way.

But there are also some disadvantages of having the future so carefully mapped out. For one, it stifles creativity. Why bother trying to think "outside the box" when all you need to do is read the Roadmap to see what challenges need to be solved. And, speaking of challenges, isn't it depressing to see such a "sea of red" in the most recent Roadmap that indicates "solutions unknown"? It's all just a bunch of extrapolations anyway.

Finally, if Dr. Smith is right, the roadmappers may simply be wrong about which technologies to focus on and which to ignore. The danger here is that, once a technology is considered unworkable for whatever reason, almost all activity stops. Focusing limited resources on the most promising technologies is one thing; completely forgetting about other options is another.

The ITRS has been an incredibly useful tool since it was introduced, and I salute all those who have worked on it so tirelessly. But perhaps it's time to set the Roadmap aside and take more of a "greenhouse" approach to solving the challenges ahead ... nurture everything as best we can and see what grows best.