Issue

Is the Roadmap losing its effectiveness

02/01/1999

Robert Haavind

Is the Roadmap losing its effectiveness?

Recently, we heard a comment from someone in the industry suggesting that the SIA`s National Technology Roadmap had become irrelevant. The complaint was that some features were shrinking well ahead of the 1997 Roadmap timetable, and that the projections for optical lithography and next-generation lithography had proven to be way off the mark (see more about this in "Optical lithography to 2000 and beyond," p. 31).

Depending on the Roadmap as an inviolate prognosticator of the future could surely lead to some misspent capital (as some developers of 300-mm tools might attest). But the death knell for optical lithography has been sounded so often in the past that an experienced observer should be wary of any predictions regarding the necessity for introducing next-generation litho. While the intent of the Semiconductor Industry Association was to provide an open, shared vision of future technology goals for everyone involved, I doubt that anyone ever felt that its predictions were sacrosanct. Variations in timetables, dimensions, and technology have always been a part of the process and have led in recent times to more frequent updating of the projections for future targets. To ensure that predictions are in tune with global, not just US, developments, the next Roadmap update will include industry participants from all over the world.

There are many more aspects of the Roadmap process that need to be considered before anyone concludes that it is an empty exercise.

The consumer electronics arena provides many examples of the folly of different companies or groups plunging off in their own directions, trying to crush standards being set by others in the industry. The fiasco of Sony`s Betamax vs. a consortium of companies backing VHS leaps to mind. Currently, multiple standards for DVD (digital video disk) are slowing acceptance in that market. A recent trial with smart cards in the US required commercial establishments to get two card readers for cards based on multiple standards. Some stores refused to cooperate even when the readers were provided free. The test was considered a failure. In the semiconductor processing industry, many years of extra expense and trouble resulted from lack of agreement on a notch or flat for indexing wafers. So one benefit of getting the semiconductor community to thrash through such issues is that the Roadmap can help define some parameters that can become standards, whether for cassettes, cluster tools, or other areas where specification details rather than different technologies are involved.

Another great value is the process itself. Dozens of experts from companies, research labs, government agencies and so on break into groups to examine future problems and potential solutions to keep the industry on the track that Gordon Moore laid out more than 30 years ago. Many views are expressed and debated along the way, and reaching consensus can be difficult. But this high-level interchange illuminates the road ahead, and gets a lot of smart people thinking about what needs to be done to make the trip go more smoothly. Surely many potential show-stoppers never happened because the alert went out ahead of time to those clever enough to find solutions.

The grumbling about how the industry is running ahead of the Roadmap timetable, and stretching the limits of optical lithography beyond any reasonable expectations, is telling in itself. Both are remarkable achievements, reflecting the energy and ingenuity of the greatest industry the world has ever known.

These are not causes for complaint. They are achievements to celebrate. Appreciate the Roadmap, and take it for what it is. But don`t blindly bet on it.