Cymer's DSM labors may finally pay off

Electronic News, Feb 23, 1998 by Dylan McGrath

10 Years Of Excimer Laser R&D But What Technology Lies Ahead?

San Diego--It may have taken longer than its founders had hoped, but one San Diego company is finally reaping the benefits of years of hard work. After 10 years of development, Cymer, Inc.--which produces excimer laser light sources for use in deep-submicron (DSM) wafer steppers and scanners--is finally basking in the light at the end of the tunnel.

"From about 1986 to 1995, we were enduring a very lengthy process development phase," said Robert Akins, Cymer president and CEO. "We originally anticipated that excimers would be the light source of choice for lithography, starting at about 0.5-micron dimensions. We were very incorrect in that assumption, because I-line mercury bulbs were extended to 0.5-micron and substantially below. The entry point for excimers got pushed off to the 0.25-micron dimension. We've had a long history; almost 10 years of technology development and market developmental positioning before the real opportunity began."

But the real opportunity has certainly come knocking. Cymer's $203 million in 1997 revenue represented a 213 percent increase over the company's $65 million in 1996 revenue. In 1995, the company had just $19 million in revenue. In two years, its revenue, and seemingly its importance, have grown by 10x.

In the early 1980s, Mr. Akins and fellow founder Rick Sandstrom, now Cymer's VP of advanced research, were part of a team at work on an excimer laser for use in the "Star Wars" strategic defense program. In the mid-1980s, convinced that DUV would triumph over X-ray technology for the DSM era, Messrs. Akins and Sandstrom decided to develop and apply their laser technology in the area of microlithography. For the next 10 years, while the company waited for the semiconductor world to be ready for excimer light sources, Cymer further developed the technology to make sure that it would be ready for the semiconductor world.

The Decisive Moment

"For a technology development standpoint and a market share garnering standpoint, we started the company at the right time," Mr. Akins said. "It took a long time to get a robust technology. Because we were early and could play in all the process development phases, we were able to get very high market share while the market was still small. From a financial standpoint, it would have been easier to start the company in 1990 or 1992 and not have as many years of no growth to contend with. If we had waited that long, some other company would have positioned itself as the company of choice and it would have been very difficult for us to come in and change the dynamics of the industry."

Cymer's 5000 series of excimer lasers are used in wafer steppers and scanners for DSM processes. Cymer's customer list begins and ends with the world's four major stepper/scanner-producing companies: Nikon, Canon, ASM Lithography and SVG Lithography. As semiconductor companies increasingly shrink their chips in order to get the most out of each wafer, DSM processes are entering their heyday. Companies need DSM steppers and scanners, and DSM steppers and scanners need excimer lasers.

Cymer, according to Pascal Didier, senior VP of worldwide customer operations, commands an 88-95 percent share of the worldwide excimer laser market. Although Komatsu Ltd. and Lambda Physik also produce excimer lasers, Cymer isn't exactly feeling the heat. "Do we have competition?" Mr. Didier asked. "Yes. But is it a competition in mass production? No. I think we have a 100 percent production share."

In some ways, Cymer must cash in on its 10 years of technology development while it can. DSM lithography's reign as the cutting-edge champion could be approximately 10 years. Crucial industry debate is under way about which technology will replace optical lithography as the state of the art. Several candidates have emerged as possibilities, but none use excimer lasers as the light source. In fact, Mr. Akins concedes excimer lasers will not be incorporated into steppers and scanners beyond optical lithography.

Does that mean Cymer is near its end? Mr. Akins points out that, while optical lithography will eventually be replaced at the cutting edge, many non-critical layers will continue to need photolithography. And photolithography will continue to be powered by the excimer laser. Mr. Akins even hints that a new product may be on the way from Cymer, one that would be targeted at post-optical lithography. "We have been very vague on the subject and will continue to be so, but we are in the design concept phase of 13.2-nanometer EUV (extreme ultraviolet), 'soft' X-ray light source," he said. "We think there is an excellent chance there may be a big market for that."

Of course, things haven't all been fun and games for Cymer. In September 1997, reports began circulating about Cymer when word surfaced that a customer had pushed out on order. Analysts and insiders alike wondered whether a flaw in Cymer's technology had been spotted. The rumor mills heated up. "As a new public company, we didn't get out in front of a few events with enough guiding information, enough facts," Mr. Akins said. "We let the marketplace speculate a little too long and the rumors took on a life of their own. We weren't out in front leading the market with facts, so they created some of their own."