Rudolph Technologies: Yielding profits

Metrology specialist Rudolph Technologies yields to no one when it comes to yield-improvement tools.

by Drew Wilson -- Movers and Shakers, 8/15/2003

YIELD EQUALS EARNINGS in the semiconductor business. So profits hinge on optimizing production yields. It's a simple relationship, but the dynamics of new chipmaking technologies are making it complex, and that's where Rudolph Technologies has been quick on the innovation front.

Rudolph supplies metrology tools that essentially improve yield. More specifically, the company has created innovations in process-control equipment for thin-film measurement and macro defect inspection.

Chipmakers who are chasing Moore's Law, investing in the required die shrinks and adopting new materials and wafer sizes, are key customers for Rudolph. A main reason is cost. One 300 mm wafer costs tens of thousands of dollars, significantly more than a 200 mm wafer. Even a small improvement in overall yield delivers a large financial benefit. Or, look at it another way: "A process mistake is far more costly today than a few years ago," says CEO Paul McLaughlin.

Paul McLaughlin, CEO, Rudolph Technologies

Rudolph's tools measure the thickness and quality of various types of films involved in the 500-some process steps required to turn a plain silicon wafer into functioning state-of-the-art computer chips. The measurements enable process engineers to quickly find and fix glitches.

Rob Loiterman, senior vice president for technology development, says the tools have proven valuable to companies trying to work with new copper interconnects. "Copper has more layers, requires supporting material, and is more difficult to work with than aluminum," he says.

For copper metrology, Rudolph has been supplying its latest generation metal-film measurement tool suited for 90 nm and smaller technology nodes. The tools are capable of nondestructively measuring up to six metal levels at high speeds.

McLaughlin adds that more than 75 percent of all the IDMs (integrated device manufacturers) that have copper processes are using Rudolph equipment.

Last year, Rudolph entered the competitive wafer-defect inspection area through the acquisition of ISOA Inc. ISOA now forms Rudolph's yield-metrology group. The group provides a system called WaferView, which allows chip manufacturers to fully automate macro defect inspection, Loiterman says.

In the past, fab technicians traditionally sat inside the factory in front of a microscope every day, identifying and classifying defects as the wafers passed by. "WaferView enables fabs to remove the operator from that step," Loiterman says.

Typically a fab uses six to 10 Rudolph tools. In a 300 mm format, which can run as high as $1.5 million for each tool, the fab can see a payback in six to nine months, according to the company.

German immigrant Otto Rudolph founded Rudolph in 1940 to develop research instruments such as ellipsometers. For five decades Rudolph Research supplied instruments to a wide base of universities and laboratories, including the venerable Bell Labs.

McLaughlin, who has 20 years of experience in semiconductor capital equipment, including six years as vice president of analytical instrumentation company Perkin-Elmer, looked at Rudolph and saw opportunity. In 1996 he led a group of investors in a leveraged buyout of the company. McLaughlin shed any product that wasn't connected to high-volume semiconductor manufacturing and changed the name to Rudolph Technologies.

Rudolph today has the largest installed base of ellipsometers in the world, a position that helps drive global brand recognition. "Anybody who ever learned anything about thin films from an ellipsometry measurement point of view probably learned it on a Rudolph tool," he says.

Rudolph's success is reflected in its balance sheet. Since becoming a public company in 1999, it has outgrown the semiconductor industry every year. But ultimately, sales are tied to capital-equipment spending, which took a dip in 2002. Rudolph's revenues also declined almost 30 percent, from $79 million the previous year to $57 million. Pro forma net income for the full year was $2.5 million.

Still, the company performed well in what was widely seen as a bottoming-out year. "Profits came down, but the company was profitable at all times and has been in each quarter through the downturn," McLaughlin says, adding that metrology is growing faster than the overall semiconductor capital-equipment industry.

Rudolph's challenge is to stay close to the customer's development strategy. "If he goes into silicon-on-insulator, copper, low-K, we must be close enough to material science to develop tools to meet his next-generation requirements," McLaughlin says.

Rudolph's chief competitor is the yield-management giant KLA Tencor. McLaughlin calls KLA a fine competitor, yet believes his company differentiates itself more on technical expertise and less on pricing. Rudolph, he adds, invested 20 percent of revenue into R&D in 2002. "We have best-in-class products in a number of areas and strong positions in market niches," he says.

McLaughlin cites several factors that he believes have put Rudolph ahead of its competitors. The company's early expertise in ellipsometry provided the platform for current instruments to measure the thinnest of films with the best repeatability. "That's the kind of expertise that we thought would be needed as things started to shrink," McLaughlin says. "The market was coming toward Rudolph in ellipsometry."

The company has also made a big splash in optical acoustics, where McLaughlin says Rudolph has 90 percent marketshare. Rudolph holds exclusive patent rights to a technology that allows engineers to use light to see through opaque metal films nondestructively and at high speed.

Another area, knowledge-based algorithms for macro defect inspection, comes from the 2002 acquisition of ISOA. "We are now well positioned in that attractively growing market niche," McLaughlin says.

McLaughlin believes Rudolph has positioned itself snugly in the metrology industry. Would-be competitors face barriers to entry such as long-term customer relationships, brand recognition, and a global support team. Many companies are unequipped for those challenges because they operate in small niches, McLaughlin says.

"Many of these smaller companies will have to change their business model if they do not get acquired," he says. "That may force smaller companies who do not get consolidated to move into an R&D business model, which doesn't require the same kind of commitment."