August - 2000 - issue > Cover Feature

Veteran entrepreneur Naren Gupta says he first met Y.C. (Buno) Pati at Harvard University where Pati briefly taught electrical engineering. In Pati, Gupta saw a young man with a personality that “excites people” to work with him. That wasn’t the only thing that attracted the likes of Gupta to Pati’s startup. Gupta, one of the earliest entrepreneurs in Silicon Valley, pumped money into Numerical Technologies four years ago also because he was convinced the technology Pati took out of a Darpa-funded research project at Stanford would have a “major impact” on semiconductor manufacturing.

Not only has Gupta been proven right, his expectations appear to have been a tad modest. Numerical has given Moore’s Law – a theory that the number of transistors on an integrated circuit doubles every 18 months – a breath of life just when it appeared to be at greater threat than ever before. What’s more, the startup’s technology is seen as ensuring the unhindered operation of Moore’s Law (see stories alongside) for at least a decade, if not more.

Pati, born in Washington, grew up and studied in Bhubaneswar (Orissa) and Ajmer (Rajasthan) before coming to the US to earn a college degree. Son of a University of Maryland professor, Pati earned up undergraduate and a doctorate degree in electrical engineering at the same university before joining Stanford University in 1992 as a research associate.

Although he had specialized in wireless communications, Pati was trying to apply signal-processing techniques to semiconductor processing for the Darpa project. In particular, the research group improved upon a 20-year-old technology called phase shifting masking, which had been applied with only limited success so far. “The first time I saw it, I was blown away,” said Dan Hutcheson, an analyst at San Jose’s VLSI Research. But in 1995, when Pati, together with Stanford professor Thomas Kailath and another researcher, licensed the technology, the challenge was to convince the market that there was a need for the technology, says Pati.

“It was like selling a new form of aspirin. People just said, ‘No thanks, I don’t have a headache,’” says Pati. Current technologies still permitted the smooth operation of Moore’s law, and the market didn’t immediately recognize a future need. Besides, the industry anticipated a new generation of chip-processing equipment based on short-wavelength lasers. Getting that mindshare in the market was the single most challenging task, says Pati. Now, Numerical’s technology is the farthest in the area of subwavelength optical lithography.

It’s a few months since Numerical started life as a publicly traded company, but Pati is not weighed down by heavy expectations. He comes to work in blue jeans, T-shirt and a golden retriever in tow. He has held a carefree approach to entrepreneurship, in spite of the fact that this was a first venture. “What risk?” he asks. “There was only uncertainty.”

Regardless of the light-hearted manner in which Pati looks back on the entrepreneurial effort, Numerical was bootstrapped at the outset. But the benefits of Stanford connections manifested themselves soon enough, and the startup developed rather easily, even though not as quickly as Pati may have wanted it to. Numerical attracted investments from Mohr Davidow Ventures, Goldman Sachs and even chip giant Intel.

“Getting a good team together was still the major task and once we had that, it was rather easy,” says Pati, who selected an experienced senior management team, some of whom had been associated with other startups.

Pati took the company public last April, when it received a warm welcome from investors despite adverse market conditions. It is trading at about $52, giving the company a valuation of almost $1.5 billion. Its roster of clients includes almost every major semiconductor player, including Lucent and Motorola. Although it’s all come together nicely, analysts believe there still remains a lot of upside potential, because Numerical possesses the most advanced technology in subwavelength lithography. Says Pati, “We are still the only company that actually produces chips at 0.1 micron level. In the lab, we have reached the 0.025 micron level.”

“For a first-time CEO, Buno is blending technology, marketing savvy and charisma,” says Rob Chaplinsky, a venture capitalist at Mohr Davidow Ventures, an investor in Numerical. “It’s rare that you get all of those in one person.”