John Heaton, Nanometrics CEO

-- Semiconductor International, 6/1/2001

John Heaton (Source: Nanometrics)

John Heaton joined Nanometrics Inc. (Milpitas, Calif.) in September 1990 and was elected vice president of engineering in April 1994. In July 1995, he was appointed to the board of directors and became general manager. He was elected president of the company in May 1996 and became CEO in April 1998. Prior to joining Nanometrics, he served in various technical roles at National Semiconductor from 1978 to 1990.

SI: This has been an exciting year for Nanometrics. You announced revenues for fourth-quarter 2000 of $19M — 52% above the same period in 1999. You also finished 2000 with revenues up 105% over the previous year. What else is new?

Heaton: The year 2000 was a landmark year for us because we broke the $40M threshold that many small companies have a difficult time transitioning through. This wasn't strictly a Nanometrics phenomenon, though, since 2000 was a year of unprecedented growth for everyone — process and metrology tool manufacturers — and the driver that allowed us to reach this level. However, in the past we've not been in the position where we actually realized the industry's growth rates.

SI: Why was this?

Heaton: In 2000, for the first time, we attained a product position that allowed us to grow with the rest of our industry. We now have a mature, varied product family that allows us to compete with all the players in our space and develop with — and sometimes exceed — our competitors' growth rate. We've positioned ourselves in different markets as the integrated solution.

SI: What are your short- and long-term plans and strategies? Will you be making changes, organizing things differently?

Heaton: Changes are under way. Last year, when we doubled our size, many of our shortcomings, vis- à -vis the way the company was fundamentally structured, became very obvious to us. We had the structure of a $30M company, and when we got to the mid-$70M range we discovered that our infrastructure, management depth, and our supply chain were constraining our growth more than the products themselves — this was not a good situation. Right after our offering — which we completed in the first quarter — we began addressing these fundamental limitations in our company.

SI: Was this move prompted by the new capital available?

Heaton: No, but it certainly helped! We generated about $75M in cash, and now that we had it, we immediately went to work resolving these long-term issues. Chief among these was whether we had sufficient physical space to house a much larger organization. At that time we didn't, so we purchased a new corporate headquarters in Milpitas, quadrupling the space that we had in Sunnyvale. We also initiated a plan to double our facility in Japan, and did additional work in South Korea because we had done so well there during 2000, and realized the value of the worldwide manufacturing strategy that led to the building of a factory there.

SI: What was wrong with your supply chain?

Heaton: Nothing was wrong as such. However, one of the issues we ran into is that we had greater demand than we could actually meet for especially integrated products. Since then, our major supplier has been acquired by Newport, and they have reassured us that they will be able to meet our expectations. Basically, it was critical for us to get additional physical space and a better capitalized and managed supply chain than we had before.

SI: What other measures have you established?

Heaton: We have also reinitiated technology development activities that have been lacking for far too many years before this. We've set up a new goal of investing 14% of our revenues on engineering. Over the last year we've doubled the size of our R&D engineering group, and are now meeting percentages more befitting to a technology-based company.

SI: Anything else?

Heaton: Actually, there's a strategy that we started over the last eight months — internalization. One of the things that we realized in the manufacture of our lower-cost metrology products is that integrated products, in general, cost less than a stand-alone metrology system. Therefore, the budget allowed for manufacturing the tool is shrinking because we aren't going to be selling a $400,000 or $500,000 tool, but selling something in the order of $100,000. Thus, development costs have become much more critical, and little dollars are more important. This is why we've started a program we're implementing in our new facility, which is to outsource less and manufacture in-house more of the assembly.

SI: So you're taking up a larger share of your own manufacturing. Why?

Heaton: The reason that Nanometrics grew up to the year 2000 is that we had an outsourcing strategy. Since then, we've learned that this has its shortcomings because, residing as we do in Silicon Valley with most of the other large equipment companies, we found that, as we grew, we were constrained. The area machine shops that make machine parts or sheet metal supplied our larger competitors as well, and we were having major problems with material delivery and lead times. This is why we brought in-house the critical components that might keep us from quickly addressing customer needs.

SI: This move cannot be inexpensive.

Heaton: Absolutely not! It's going to require greater capitalization because we're going to hire more people to do some of the lower-level work, and buy manufacturing equipment. We're going to manufacture, design, and build and assemble our key system components and technology in-house.

SI: In view of all this, what do you see as your next growth areas?

Heaton: First of all, we're addressing infrastructure and our ability to manufacture our own products. Secondly, we're addressing the R&D investment. Now that we have a platform that is integrated onto process tools, all we have to do is find more metrology subsystems that can be designed into this platform which is already being designed into products, in order to increase the level of acceptability of the products and broaden the addressed markets.

Today, Nanometrics ships on CMP and CVD tools, but we believe that there are many other metrology systems that can be incorporated onto our current platform, that would expand it far beyond just CMP and CVD film thickness. We think that scatterometry is a key technology for the industry, and intend being involved in it by putting this subsystem into our platform.

There are other dimensional metrologies, such as overlay, in which we already participate, as well as topographical measurements for CMP dishing and erosion that would be very compatible with our system and technology. These are all opportunities for us, and if we can execute on the platform, we'll have a much broader market for the company.

SI: Obviously, you believe the demand is there.

Heaton: No question! Over the next five years, as the shift from 200 to 300 mm progresses, wafers will become much more valuable for the customer, increasing his risk. So as a risk-reduction and productivity-gain strategy, we believe that, now that metrology is starting to move into the process tools, its benefits are becoming increasingly obvious. Most companies executing 300 mm pilot operations are incorporating integrated metrology.

SI: Do you think growth is getting more difficult to achieve than before?

Heaton: No. Over the past few years metrology systems have become more important because they provide the benefit of monitoring the process. Another thing we see with metrology in general is that prices are rising, not falling.

In the process tool area, all the big players are struggling to put out very high-throughput systems and undercut each other's prices. In metrology, however, the opposite is true. Systems aren't necessarily getting much faster — they're getting larger and more complicated and more expensive.

So, from a revenue viewpoint, metrology is a growing area, and we expect it to outgrow the overall equipment space. On top of which, during this transition, we're going to continue to see fabs spending on stand-alone metrology as well as in the integrated metrology area. The metrology budget for most facilities will continue to increase.

SI: Obviously, metrology is no longer viewed as a non-added-value expense.

Heaton: Definitely! This is driven by the fact there are few areas for improvement in fabs today. Metrology made a huge impact in the 1990s, when we got yields to 95% and above. Now we're finding that metrology not only can make a difference in this extra 5% for the fab, but there is also the transition to 0.13 µ m processes — people are realizing that, unless they're able to monitor processes more closely, it becomes increasingly more difficult to keep them under control.

Before, 1.0, 0.5 or even 0.35 µ m processes were fairly straightforward and didn't require much monitoring. At the 0.13 µ m node the process doesn't remain under control as easily as it did in the past. Manufacturers will need to monitor more closely because processes are inherently more unstable — photoresist is touchy, films are getting thinner, the materials and geometries we're trying to deal with are much more critical. Any excursion could prove disastrous to productivity, particularly with larger wafers containing as much as 2.5 times more die.

SI: Increasingly, tool manufacturers are asked to produce equipment that must meet more stringent requirements at lower prices. Can anyone alone foot the research and development costs involved in the move to 300 mm, copper, and all the rest?

Heaton: Certainly each customer — particularly in the 300 mm area — has a slightly different concept of how he wants to implement 300 mm in the fab. It seems that every order we get for a stand-alone system outlines much more difficult requirements than the tools of the past — each fab becomes almost a custom engineering job. Some of the standards work done by SEMI has been very useful, but there is still too much variability, particularly between countries.

Bottom line, costs are being passed on to the customer, and they are higher than before because metrology requirements are far more sophisticated for 300 mm than they ever were. The trend, as I mentioned, is toward integration, so if in the past they paid $500,000 for a film thickness tool, it is now over $1M because of all the extra requirements for door openers, extra robotics, air handling, etc. When you put the metrology at the process tool's front end, you eliminate the cost of all those extra requirements.

SI: Customers now not only want vendors to provide better and cheaper tools, but also the expertise that they lack in-house. How do you view this trend?

Heaton: Over the past few years, to be a player you've had to provide direct applications and service support where the customers are. So the expense associated with creating such a worldwide support network has been growing. However, if you want to play the game, that's what you must do. Obviously, there are critical-mass considerations as to whether a company is able to do it, but I don't see it as a major issue — it's an expense of doing business.

SI: Are you satisfied with the level of information exchange with your customers, or do you wish they were more definite about what they want and why?

Heaton: If you're working closely with a customer, this generally establishes a long-term relationship. After you have a product in a fab, most customers make it very clear what it is they want and why. Also, when you work in the fab with your customer, his needs become obvious. Another factor is that the amount of variability that used to exist between customers has come down — most everyone is singing from the same hymn book as far as metrology is concerned.

SI: How do you view the trend toward the fabless model?

Heaton: Some are concerned about it putting too much control in the foundries' hands. Obviously, it hasn't worked out that way. The foundry model is a very strong influence in our industry. Fabless companies must make a considerable investment in foundries, and we're seeing some shortcomings now because fabless companies don't have the end markets when times are bad, and foundries suffer disproportionately.

If you look at IBM and other traditional, large semiconductor houses, they're always putting out product regardless of demand. Samsung, for example, stamps out the same number of DRAMs month after month, regardless of how the market prices them. In the case of the foundries, when the customer says he doesn't want any more chips, they don't have anything with which to backfill. During bad times, foundries suffer more than other companies with their own built-in facilities.

SI: Is there too much consolidation taking place in the industry?

Heaton: In metrology, I don't think there has been that much consolidation taking place. I believe that with the new accounting standards this will accelerate, however. If the downturn continues, some of the smaller companies will be taken over — they'll begin to suffer and turn toward larger ones for help, which means sale or merger.

SI: You mentioned the current downturn. What, if anything, could we do to reduce or eliminate these industry cycles?

Heaton: I prefer us having these cycles. They allow the good companies to reposition themselves and penalize the bad ones. Thus, when capacity buys come back on-line, the good companies with the long-term view, which have worked hard during the downturn, benefit. For us, it's a great time to retool for greater growth and reposition our products for the next upturn.

SI: What are the industry trends we ought to be paying more attention to?

Heaton: On the process tool side, the fact systems are getting more demanding because they are faster and more efficient. The presence of large, dominant companies such as Applied Materials helps because they establish standards everybody can work with.

SI: What do you see as some of the industry's ongoing problems, and how would you change them?

Heaton: Equipment manufacturing here in Silicon Valley is a major problem for us. During the good times, all tool companies have problems because we're all trying to use the same facilities for machining and parts construction. Then, when Applied Materials or KLA-Tencor get going, there's little capacity left for the rest. This makes it difficult to execute as quickly as customers would like.

I would like to see more globalization in equipment manufacturing — I would like to see much of it leave the Valley and spread ... ( smiling ) for it to go somewhere where they don't turn your power off when you need it most.

SI: From your viewpoint, which are the biggest hurdles in transition to copper?

Heaton: The big problem in copper today is in CMP. It's a difficult process, and there are very few companies that have worked it out. I think there are problems, with today's technology, in the erosion and dishing areas — people are unable, in a non-contact, production-worthy method, to monitor or measure these factors. This hurts everybody's capability to characterize processes faster and more accurately.

SI: What will Nanometrics look like in 2006?

Heaton: We'd like to be a half-a-billion-dollar company — we will be regarded as the leader in integrated metrology with 50% market share or better. We will have multiple product offerings for all the compelling problems that process tools have today, and we will provide good value for our investors.

— Alexander E. Braun