Executive Insight: CH Wu

An insider’s look at doing business in Taiwan, the growing market in China, and how that affects the semiconductor, foundry and test markets.


Semiconductor Engineering sat down with CH Wu, president and CEO of Advantest Taiwan, to talk about business, politics, and his philosophy on what really motivates people. What follows are excerpts of that conversation.

SE: Tell us a little about who you are and your background.
Wu: I graduated from college with a degree in electrical engineering and started at Philips Electric, then moved to Motorola, which today is Freescale. I’ve been working in semiconductors almost 30 years now. I’ve been at Advantest for more than 20 years. I’ve had many jobs, from service to engineering and even sales. Doing all of those roles helped prepare me to manage this company today.

SE: Over the past three decades you must have seen a lot of changes in a very volatile industry. How do you keep your company and yourself focused and positive in that kind of environment?
Wu: If you are in the industry long enough, then you can see the cycles and prepare for them. Those cycles do vary over time. In the past, one cycle might be five years or three years. Today it may be three to six months. Change is faster and faster and everyone is worried about time to market.

SE: What’s your biggest challenge for 2014?
Wu: The market itself. Test is still growing, but not like in the past. Ten years ago it may have been growing 10% a year. Now it may be closer to 3% to 5%. But there’s a lot of new technology, and that technology generates a lot of data. The challenge is how we bring the technology from Germany and Japan and deeply engage with our customers and extend the technology they can engage with. Also for Advantest over the past 10 to 20 years in Taiwan, everyone knew us as the memory company. Today we have more than 5,000 or 6,000 systems installed. After we acquired Verigy, we needed to shift our focus to SoCs. Today, we are providing a total solution. We are not just addressing memory or even just SoCs.

SE: Did the acquisition of Verigy help change that perception?
Wu: I believe so. From my perspective that was the right decision for Advantest and we’ve done well with it.

SE: Advantest started as an instrumentation company. How has it changed over the past couple decades?
Wu: Today, addressing the test challenges of SoCs is a major business for Advantest. Of course, memory is still growing, but the customer base, through consolidation in recent years, is getting smaller and smaller. There are three major companies and the customer base is shrinking. But at the same time the volume is growing, so this is still a big market. In the past, this market was driven by big data, but all the computers in home electronics need memory, as well.

SE: How does Advantest differentiate itself from its competitors?
Wu: My personal opinion is this has to be done as a partnership with customers. We’re not just sales assistants. Without that partnership you don’t understand the benefits for each side. If you’re just selling the machine, it doesn’t work long term. A partnership approach does work, and customers definitely appreciate it.

SE: What are some of the biggest challenges faced by the ATE market as a whole?
Wu: The growth rate may never be as large as it was in the past. That’s why we pushed into a total solution approach. We also moved into MEMS and other applications. But for ATE to continue growing, we need to enlarge the market to fit different applications.

SE: We’re seeing a lot of consolidation in ATE—Advantest and Verigy, LTX and Credence, Teradyne and LitePoint. There are some smaller companies providing test in Korea, too. Do you see more consolidation ahead? And is that even possible?
Wu: There are many small companies in Taiwan. They try to copy our approach, but they can’t achieve the economies of scale necessary to provide a total solution. Still, I don’t think there will be much more consolidation.

SE: If you look at front-end equipment, we’ve seen consolidation there, too—Varian, TEL and Applied Materials, Lam Research and Novellus. Do you see more consolidation ahead there?
Wu: It could happen. There are still a lot of companies left in the front end. My opinion is that it’s more difficult. It will happen, but it won’t be easy.

SE: Where do you see the best options for growth?
Wu: In Taiwan, the foundries and design houses and OSATs are well developed. That’s different for China, which will be the next major market for semiconductors. That will be a substantial change. As you know, in the past five years the Chinese government has done work on the foundry side, and I believe in the next 10 years they will build their industry on the high end, like CSP (chip-scale package) or BGA (ball-grid array). In Taiwan, our customers are strong on the OSAT side. In my opinion, these customers will need to engage more heavily in China.

SE: Many companies in ATE and semiconductors have had to merge with other companies to remain strong in a country. As China emerges as the next area of growth, if you’re not doing business in that country is there an opportunity you’ll miss it?
Wu: Yes, I believe so. There is a lot of growth opportunity in that region.

SE: How has the Taiwanese semiconductor industry changed over the past decade? OSATs have always been a big part of the market, and there are large foundries.
Wu: The Taiwanese semiconductor industry has been in development for the past 30 years. The last 10 years haven’t seen very big changes, though. TSMC has taken the leadership in the foundries. ASE is still the winner on the OSAT/assembly side. In Taiwan, this industry won’t be the center of the IDM model. The focus is on the OSAT model. But we built a very good supply chain for semiconductors. In the past 10 years, those companies developed advanced technology all the way down to the 1x nanometer generation and provide services to the customer. The next 10 years will be a bigger challenge with 3D-IC and 2.5D, because so far no one has taken ownership. The yield and process for TSVs is still not stable. TSVs, packaging and yield must improve. Where are the defects coming from? Is it the wafer side or the packaging side? It’s very difficult to judge. The cost is still too high and the yield isn’t there, but the bigger problem is who’s going to take ownership of that approach. That isn’t defined yet. On the assembly side, ASE provides a 2.5D solution.

SE: You spoke about that at a GSA event a few months ago. Has there been any progress since then?
Wu: There has been some progress. But for the 3D-ICs or SoC stack with memory, it has been slow. For SoC-to-SoC, they can do that now. TSMC can do that alone. For SoC-to-memory, that’s more difficult.

SE: One question we always ask CEOs is what keeps you awake at night?
Wu: Many of my old friends are retired already. For me, I wouldn’t know what to do. I enjoy the work. But the question you always have to ask yourself is who you are and what value do you bring to your company? From the overall market side, over the past 10 years semiconductor technology has brought huge benefits to people’s lives. But we may be consuming too much. We have two cars and two cell phones. In the long term is that good for us, our children and our environment?

SE: If you had one message to leave readers with, what would it be?
Wu: Again, you need to know who you are, what value you provide and why you are doing this. Is it for your family or your colleagues or your company? If you have a strong mind and you’re determined, you can do better. No one can force you to learn or to work unless you’re self-motivated. You need to know who you are and what you want, and that can help you to move forward.

Leave a Reply

(Note: This name will be displayed publicly)