Wen-chi Chen: Taiwan's Answer to Andy Grove?

Via Technologies' ambitious CEO plans to win by trying to play a different type of game than Intel

By Bruce Einhorn

Wen-chi Chen, president and CEO of Taiwan's Via Technologies, has launched an audacious bid to transform his company. Without a single chip-fabrication plant of its own, Via has grown into a $1 billion company by contracting its manufacturing out to foundries.

Thanks to some stumbles last year by Intel, Via is the world's No. 1 maker of chipsets, the semiconductors that connect a computer's microprocessor to its graphics, audio, and other functions. Via is now trying to market its own microprocessors. And, in a first for Via, the CEO has discussed setting up chip facilities in China. Chen talked to BusinessWeek's Bruce Einhorn on Feb. 6 at a chipset-technology conference where he was the keynote speaker. Here are edited excerpts of their conversation:

Q: Why are you pushing Via into these microprocessors?


We have to. If we didn't...we wouldn't be able to compete. I don't think the chipset [market] is going to be the same. In the future, the chipset and the graphics and the processor will integrate. It's the natural thing. The chipset is the centerpiece, and it will continue to integrate with a lot of things. We've added this and that over the years. Today, we are adding graphics and audio into [the chipset]. In the future, we will add Ethernet [network] controlling, peripheral controlling [functions]. We may need to add the processor on it.

Q: Nobody else from Taiwan has really succeeded in breaking out into the consumer market, as you are now trying to do. Acer, for instance, is the biggest computer maker in Taiwan and sells a lot of PCs to U.S. companies as a contract manufacturer. But it hasn't been able to make a dent in the U.S. consumer market. What makes Via different?


We and Acer are quite different. Acer is trying to be a market follower. I don't [think] they're a technology company. Via is more of a tech company -- we spend a lot of effort to develop tech and promote it globally.

Q: In your speech, you made a couple of jabs at what you called "the 'I' company." Do you enjoy seeing yourself as David to Intel's Goliath?


I am a Christian myself. I don't like to classify Intel as a Goliath. I would like them to be part of the David camp also.

Q: You were able to increase your market share in chipsets last year to over 40%, leaping ahead of Intel, thanks to problems with the Intel chipset that used technology developed by U.S. startup Rambus. Looking back, where do you think Intel went wrong?


Intel unfortunately made a big mistake, and that cost them a lot and cost the whole industry a lot. We advised them way back in 1998 that this could be risky.... We told them that [we] thought there were quite serious problems with that technology. We pointed that out, and they didn't listen to us. But when we talked to some [of our] other partners, they said that they had similar concerns.

Q: Now you are taking on Intel in its core business, microprocessors. What do you think gives you an edge over Intel and industry No. 2 Advanced Micro Devices?


Compared to Intel and AMD, we are focused more on the value line. We focus on cost, performance, and low power consumption.

Q: But lots of companies have tried to make a dent in the dominance of Intel over the years. Other than AMD, they have all fallen into obscurity. What makes you think Via will be able to succeed where others have failed?


The others were playing the same kind of game as Intel. The analogy for this is an NBA basketball game. Intel is like 7'2" tall. AMD is 6'11". But we're trying to play a different type of game. Even though we're shorter, we're fast. And it's easy for us to move around. That gives us a better advantage. Because we're fabless, we have very light infrastructure costs. We have a very strong position in Asian sites. And we're also focusing our product more [on] the value-line space.

Q: What's the advantage of focusing on cheap computers?


We think that the value-line product will be the fastest growing in the whole PC space. It can easily grow another 20% higher than the whole market. That's the market where people want the products. People are looking at [a] second or third PC for the home, and the value line serves that purpose better.

Q: Why?


The Internet will be the biggest driving force. For Internet applications, the processor is not the main requirement -- the pipe is more important than the engine. It's like the auto industry: It used to be that people build bigger and bigger engines. But the road was only so wide, and car drivers were more limited by the road than by the engine. So people started looking at efficiency of the engine and went to smaller engines.

Q: You're now talking about expanding into packaging and testing. But some critics contend that the move makes little sense, since you have little experience in that area, and it's far removed from your core business.


The main reason is that we have certain needs that drive the technology more. Testing is closely related to our business. Right now, we mostly use outside packaging and testing for chipsets. For microprocessors, we do all the testing in-house. But we look forward to work with [a] partner and migrating that business.

Q: So you're still committed to the fabless model?


We think that [the] fabless design house will continue to be the trend, except for memory-related companies like Micron. Those companies need fabs. But for all the others, the future will be fabless houses working with foundries. That will be the trend. We think that we can take advantage of that trend and become one of the few major players among the fabless IC [integrated circuit] design houses. Broadcom, Qualcomm, and Via -- all of us will become much bigger than before.

Q: What's your goal for the microprocessor business? Do you intend for it to become a major part of your business mix soon?


In the coming year, we think chipsets will still be the majority of our business. Now, 90% to 95% [of our revenue] is from chipsets. But we would like to be No. 1 in microprocessors, too. That will be a while, a long while.

Einhorn covers technology from Hong Kong for BusinessWeek. Follow his weekly Online Asia column, only on BW Online

Before it's here, it's on the Bloomberg Terminal. LEARN MORE