Oil Shortage May Impact Semiconductor Industry
Kirk P PondPresident, CEO & Chairman Fairchild Semiconductor USA
A spin-out from National Semiconductor in March 1997, Fairchild Semiconductor had doubled its revenues by its August 1999 IPO, and now has a phase-two growth plan to double revenues again by 2002. To achieve this, the company plans to raise its R&D expenditure to generate 70-90 new products per quarter, to increase its capital expenditures, and to continue to grow through acquisitions.
Asia played an important role in Fairchild's business, accounting for 63% of the company's US$1.3 billion revenues in 1999. Sales are expected to grow over 60% this year. About 80%, or 9,900, of its staff are stationed in Asia, where it has five manufacturing facilities in China, Korea, Malaysia and the Philippines.
Kirk P Pond, President, CEO & Chairman of Fairchild, talked to NEA about the current state of the semiconductor industry and the company's strategies.
NEA: Fairchild offers a broad portfolio of products for multimarkets. What are the main advantages of this strategy compared to a more focused approach?
Pond: There are a number of advantages. First of all, our products are building block products. For example, a power MOSFET can be used in computers, mobile phones, tape recorders, etc. So I'm not necessarily designing products for one application, but for lots of applications. The second advantage is that these products tend to have a relatively long life in the market; typically my products will last in the market place for 10-15 years.
The third, and a big advantage, is that our cost structure is much different. Instead of a US$1 billion or US$2 billion fab, my products can be built in facilities that cost a fraction of that amount. Instead of having a high fixed-cost percentage, I have a high variable cost percentage. If business is good, I can add capacity quickly; if business is bad, I can adjust my capacity quickly. Those things give me a lot of advantages to ride out the up and down of the semiconductor market. Of course, we ship products in quantities of 18-20 billion units a year, that allows me to have very competitive cost.
NEA: You plan to raise R&D expenditure to 5-7% of revenues to produce more new products. What are the focuses of these programs?
Pond: Probably about 75-80% of our R&D money goes into power and interface products. In power analog and discrete areas these would be power supply, power management, power routing, power conversion, etc; and in interface and logic area these would be interface and bus interface components. These two areas represent about 50% of our revenues. If you look at our analog and discrete businesses, about 65% is in the power area. Now we have very strong position in power analog and power discrete. Our analog business was zero two years ago, and today it's almost US$4 million. We have a very rapidly growing analog business, which is one of the best places to be in the semiconductor business.
NEA: Recently there were reports of a general shortage in supply of parts and components, how's the situation now?
Pond: Lead time has been growing for quite a while and people are worried that this is leading to a downturn. Our industry has grown very rapidly. This year the semiconductor industry is estimated to grow 35%, that is twice as high as the historical rate which was 16-17% for the last 30 years. So I think this type of abnormal growth rate is causing some shortages, but that will start level out a bit. Next year's growth rate is expected to be 20-25%, still above the average but not nearly as robust as this year. So I tend to think that a lot of these shortages were phantom shortages. The right growth rate in our industry is probably about 20%. If it gets much higher, people are probably not consuming all the components, but they are building up the inventory back at some place. So I would say 20-25% is the sustainable growth rate for a few years.
NEA: Did the shortages affect your company as a supplier?
Pond: I don't think that has affected our business. For instance, people couldn't get the tantalum capacitors for building wireless handsets, and therefore they didn't need our power MOSFETs; then I would simply ship those power MOSFETs to people making laptop computers. We have the ability to move our production around to where it is needed. I don't think there have been lot of shortages. Flash memory was reported to be a shortage. This product primarily goes into wireless handsets, and now people are adjusting their wireless handset bill rates, and so now the alleged flash memory shortage is coming down.
NEA: Apart from higher-than-usual growth rates, what are the other problems that you think the industry is facing?
Pond: For the last three or four months, financial analysts around the world have been trying to decide whether the semiconductor industry is at the beginning of a cycle, mid-way into a cycle, or at the end of a cycle. I think that is an important question because it impacts investment decision. Obviously if I thought we were in the cycle, then I would not be investing in new factories.
I think this oil shortage situation is potentially a very troublesome issue, because semiconductors are now very pervasive in the world economy. If oil shortages or oil prices impact the world economies, as they did for instance during the Gulf War, those economies were slow, and this certainly would impact the semiconductor industry.
Another thing is that we are going from a growth rate of 35-40% this year to 20-25% in 2001. That transition is very important because if our customers keep ordering products with an unrealistic growth expectation, we could be building inventory and of course that would cause us a problem. But I'm very optimistic. Short of the oil crisis causing a problem, I'm very optimistic that we are going to have a very good year next year and the year after.
NEA: Can you explain your e-business initiative?
Pond: There are several aspects. The first aspect is order management, such as order taking and order tracking. Our sales and billing systems are Internet-enabled, so people can give us orders over the Web. We also have an initiative to do design over the Web. So, for example, an engineer sitting in Shanghai can get onto the Web, look at our latest interface parts, perform simulations with the tools we have on the Web, and order samples. The third area is customer services. Customers can check the inventory position, lead time, pricing, etc. They can go back into our system and see where the orders are and when they can get them. This allows them to have more confidence in getting their orders on time. We are in the second year of this three-to-five-year program. The first part of the program was put in place, and now we are hooking these tools onto the system.
NEA: Can engineers now access the design tools you just mentioned?
Pond: Pieces of that are on the Web now; there are logic interface simulator tools and simulation tools for power MOSFETs. These capabilities will be built upon, but pieces are on the Web now. And of course we have all our databooks, datasheets of all our products. We have a customer base of 50,000, and we are introducing 300-350 new products a year; there are 15 million design wins that we would like to get, so we need a lot of electronic help.
There are one million electronics design engineers around the world doing design works right now. We want to be able to get to these people and get them to use our products, so we have to use electronic tools. We want to be the leader in Internet-enabled design works. Since we started our company 3.5 years ago, we spent about US$60 million on new software tools, and probably another US$25-30 million over next two years.
by Keith Chan, Hong Kong
(December 2000 Issue, Nikkei Electronics Asia) |
