Novellus CEO Richard Hill Looks at Growth and Competition in the Equipment Industry

-- Semiconductor International, 12/1/1999

Richard S. Hill has been CEO and a director of Novellus Systems (San Jose, Calif.) since 1993. From 1981 to 1993, he worked at Tektronix, Inc., as president of the Tektronix Development Group, vice president of the Test and Measurement Group, and president of Tektronix Components. Prior to this Hill held engineering management positions at General Electric, Motorola and Hughes Aircraft Company. In 1997, he joined the board of directors of SpeedFam International Inc.

SI: We've just weathered our industry's last downturn, and people already are talking about the next one. What do you think should be done to smooth out these cycles?

Hill: It depends on which industry sector you're talking about. In the semiconductor industry as a whole, if you look at the growth of semiconductors from a unit-volume standpoint, it is pretty linear year in and year out. Certainly because of the industry's competitiveness -- particularly the memory segment -- you see a dramatic up-and-down swing, which is largely pricing- rather than unit-volume driven. From that standpoint, supply control is the key to a less volatile market. You cannot control demand -- that's random -- but you can control supply. If the applications everyone is talking about materialize, demand will shoot up, and the capability to put out the capacity should help smooth the industry out. It won't be possible to produce fast enough to meet the demand, and, as a consequence, prices should become somewhat more stable.

If you move one step down in the food chain to the equipment market, particularly processing equipment -- front-end equipment -- producing 50 to 100 wafers an hour, the nature of the business is capacity increments. Naturally, fabs want to maximize their efficiency, and you're most efficient when you're at 100% capacity. Once this is reached, a decision must be made whether to add capacity. This triggers an incremental growth in equipment purchases, which causes an upturn in the equipment business, followed by a period of digestion. I don't believe this will ever go away because it is the nature of the business. As capital equipment suppliers we can mitigate this through some diversification of the types of products we offer; however, I think that we'll always have cycles. Do they have to be as extreme as they have been? No, but we'll always have them.

Richard S. Hill, board chairman of Novellus Systems. (Source: Novellus Systems)

SI: Novellus just went through an 'organizational restructuring.' What are your strategies and plans, now that everything is in place?

Hill: Ours is a relatively young company, and when we look at our projections for the future, we see some substantial growth opportunities. A challenge for any company undergoing the transition from small to large is management: How do you develop and empower management to run the business the way you ran it when it was $100M? Our reorganization created business units focused on existing products, and combined our manufacturing and engineering organizations in a way to develop managers within the company who will take on broader responsibilities. We're also molding the organization to better interface with our customers' business line.

SI: Do you foresee changes in the corporate culture?

Hill: [Smiling] I'm hoping the good things will remain and the bad ones be discarded. The corporate culture is the result of the agglomeration of people who are working together, and hopefully you have some overriding values that drive the organization. I believe we do.

SI: Are you planning to change anything else?

Hill: I'm always planning changes. If you operate in a certain way for six months you have to change it to keep it exciting for people. Novellus will always be changing, every single day. Most people fear change; I believe it brings opportunity.

SI: What are Novellus' next large growth areas, and how are you planning for them?

Hill: Certainly, copper's acceptance is a major opportunity for which we are well positioned. We were waving the copper flag long before many realized just how big it was going to be. Over the next five years, the whole metal interconnect will move from aluminum to copper. We are at the heart of this move. Beyond that, we see some tremendous technological challenges in the interconnect structure itself, and we believe we can bring technology and innovation that can lower costs and increase productivity.

SI: Is growth getting more difficult to do than before?

Hill: The good news about being small is it's easier to grow than when you're huge. I believe we can sustain our double-digit growth rate for the foreseeable future, but we have to execute.

SI: What do you see as some of the endemic problems afflicting our industry?

Hill: One of the biggest single problems the industry faces is the concentration of power in a single vendor. For whatever reason, as the industry matured, competition was squashed out. Maybe it's an indication of the difficulty of making money in the capital equipment business. I believe that a lack of vitality and competition in the equipment sector threatens the semiconductor industry as a whole. This is something that should concern semiconductor industry leaders.

SI: What are you planning to do differently from your competitors?

Hill: The competition operates in a way that seeks to control the industry. I don't want to control the industry; I want to respond to it. I cannot drive the semiconductor industry. Some of my larger competitors believe they can. We're a responsive alternative to that approach.

SI: Increasingly, users ask equipment manufacturers for more sophisticated capabilities at lower prices. Are the requisite R&D and platform development reaching a level of cost and effort that will require the end user to foot part of the bill?

Hill: Historically, end users footed the entire bill. The capital equipment industry was more or less a 'job shop.' Applied Materials must be credited with changing that paradigm, when they began developing and marketing equipment and offering new ideas to users. I don't see the trend going in the opposite direction. If you look at some of the bigger players today, they are capable of funding their own R&D and, in fact, have a tremendous amount of new opportunities and innovations in their R&D labs that I believe will become a 'breadbasket' from which semiconductor companies will be able to pick and choose. Undeniably, it's getting more expensive to do new equipment, and riskier for the smaller companies. However, I believe well-funded companies like us, Applied, Lam, and KLA-Tencor, which have a good cash flow and plenty of innovation, have great opportunities.

SI: Companies are trying to provide more services to their customers. Do you see something like Applied's EPIC blueprint as applicable to Novellus?

Hill: I don't espouse the EPIC principle. I believe it's pretty arrogant for a company in the equipment industry to think it can attempt to advance semiconductor technology. Just think about what goes into developing a process; it requires hundreds of thousands of wafers, advanced structures, long-term reliability testing. Those costs are amortized over huge unit-volumes of product; otherwise, you couldn't afford to do it. If you took that model and put it into an equipment company, then tried to amortize those costs over a relatively few pieces of equipment, it would be very difficult -- if not impossible -- to do. Not to mention the fact that it creates an entity that doesn't want to change.

For example, let's say that a company develops an advanced state-of-the-art process at a cost of $100M, and they deliver it to semiconductor Manufacturer A. Manufacturer A, who competes with Manufacturer B, starts production. Then, when the equipment company gives the same exact process to Manufacturer B, Manufacturer A will want a new process to compete with Manufacturer B. However, the equipment company doesn't have a vested interest in developing a new process for another $100M to aid Manufacturer A, because they must recoup the $100M they invested in the original process. This model has a built-in inertia that stifles development.

SI: What are some of the major hurdles in the transition from aluminum to copper?

Hill: The only hurdle is intellectual understanding by those we have trained in the semiconductor industry -- nothing else. Thirty years ago, when the industry was beginning to develop, aluminum was selected as the material used for the interconnect. This wasn't because of its superior electricity-conducting performance; it was a compromise reached due to its capability, vs. copper's, to be etched. Back then, the world would have loved to have copper. As process technology evolved and geometries shrank, a new material was introduced into the marketplace -- tungsten -- to be the vias between metal layers of aluminum. Again, not the most ideal for conductivity purposes, but it was able to fill very small holes. That spawned the development of CMP. With its advent came the capability to remove metals differently. CMP enabled copper, with an inlay (or dual-damascene process), technology. It's only a matter of time until engineers' knowledge, skills and abilities are as well versed in that dual-damascene process as they were of subtractive metals, at which time 100% of this business will be copper.

SI: What would you like your generic end user to do differently?

Hill: There are many things. Above all else, I wish they'd understand that, although their ultimate goal is faster, better, cheaper, there must be a balance between those variables, and that this involves a tradeoff. I don't mind the customer trying to coach me into doing a better job; but there are different kinds of coaches. Some try to stimulate you by telling you how lousy you are and how everybody else is so good. I'm not particularly well motivated by that form of stimulus. I prefer they tell me that if I can do such-and-such there is a great opportunity, regardless of how great the effort. Encourage me that if I make the jump there is a product at the end of the rainbow, but don't make a promise you can't deliver on for a product goal at the end of the rainbow.

SI: Any acquisitions on the horizon?

Hill: [Smiling] You really expect me to answer that? I'll say this: by company design and managerial philosophy, we're not out to create a conglomerate. When I analyze my competition, I find some are becoming a conglomerate and distribution channel, instead of a high-technology capital equipment company. When you begin piecing together various parts of the industry, your major asset becomes the distribution channel instead of technology and innovation. I don't intend taking Novellus down that road. However, that doesn't mean that we're not looking out for an opportunity to acquire technologies or businesses that could benefit from being with Novellus. But that's not our main strategy for growth. We look for growth from within.

SI: Do you think, then, that there's a certain critical mass beyond which doing technology and serving the customer become incompatible?

Hill: When you get too large, trying to bring along various pieces of technology simultaneously becomes extremely complex and difficult. In an industry that's made up and created by technical excellence, the dilemma we face is that as you become broader it becomes more difficult to have all the links in the chain be equally strong. There is a fundamental flaw in the strategies being proposed by some of our competitors. They'll use the capability to bundle two pieces of equipment as an overwhelming advantage to buying the superior unit process solution. Any process is as strong as its weakest link. If you have a division that isn't performing and doesn't have a competitor's technological excellence, you want the flexibility in your company to search for alternative solutions to shore up your weaknesses. I can always try to be the best in everything, but I cannot always be the best in everything, so I have to create an organization that can deal with that.

SI: What will Novellus be like in two to five years?

Hill: We'll be three, if not four, times as large as we are today. We'll be a dominant player in the deposition business for high-speed interconnects with the most cost-effective solutions in the market, and be viewed as innovators.