Once facing bankruptcy, Taiwan-based wafer contract manufacturer Win Semiconductors is now the world’s largest gallium arsenide (GaAs) foundry semiconductor company with a market capitalisation growth of 640% in eight years. Having partnered global first-tier manufacturers and mastered the technical blueprint for the next decade, how is Win planning to take it to the next level and maintain its lead?
Win Semiconductors chairman and president Dennis Chin-Tsai Chen, discusses his business philosophy over the past two decades.
Huang: Win Semiconductors was founded in 1999, and at that time, the market was dominated by global companies, and there were many newcomers and the competition was fierce. Soon after [in the early 2000s], the infamous tunnelling fraud involving chipmaker Procomp Informatics took place. In addition, Win Semiconductors was losing NT$1bn ($28.8m at contemporary rates) a year and was in a precarious situation.
In 2003, you took over the management of the company and focused on streamlining the business and ramping up research and development (R&D) efforts, which helped turn its operating loss into a profit in just a few years. Why did you make such a bold decision at that time? What if the R&D initiative did not produce any results? Please share your journey along the way.
Chen: When Win Semiconductors was founded, we were still living in the 2G era, and the whole industry was expecting 3G to arrive shortly and that there would be unlimited business opportunities. In the end, all we heard was the sound of the stairs, but no one had come down. Therefore, around 2000, many companies around the world could not survive because they could not see the future.
At that time, we had an excellent team who had returned to Taiwan from the United States, and we wanted to emulate chip contract maker Taiwan Semiconductor Manufacturing Company (TSMC)’s successful model of silicon wafer manufacturing and become a compound semiconductor foundry. After I took over, I thought we would not make it because we were all lab technicians and had no experience in mass production. At that time, the three integrated device manufacturers (IDMs) – Skyworks Solutions, TriQuint Semiconductor and RF Micro Devices [later merged with TriQuint] – had a 95% market share and were already satisfying the market needs. How could we compete as a new foundry company that had inferior technology and no mass production experience?
We had almost burned through all our money in 2003, and it seemed like we only had NT$500m ($14.4m) left – a semiconductor company with that kind of cash is really unable to operate. I, therefore, reduced all the production lines because why waste time when we could not do business? Fortunately, we found an acquaintance from a leading company in the industry who came to Taiwan on a voluntary basis and gave us a full week of lectures from the perspective of their leading position, clearly explaining the technological blueprint for the future of the industry to us. We also learnt the technology we needed at each stage, so we were no longer blind men touching an elephant, figuring out its entire shape by the trunk or the tusk.
Huang: So you got a blueprint.
Chen: Yes, and it was from a global perspective. The gentleman even formed an advisory group to support us after he went back to the US. He identified the problems according to the technology to be developed in the blueprint, and our team of PhDs worked hard in groups, and the American group would come to Taiwan every three months to check on our progress.
We were lucky – when we were at our wits’ end, we found Taiwan’s Lord Mengchang [prime minister of the ancient state of Qi 475 BC-221 BC, known for his leadership and generosity], Yeh Kuo-Yi, co-founder of original design manufacturer Inventec Corporation, who helped provide resources. And fortunately, because our technology fit the market, we were getting close to becoming an IDM in 2006.
At that time, Avago Technologies (now Nasdaq-listed semiconductor software product maker Broadcom) saw the coming of 3G and wanted to enter the wireless communication field, so they approached us in 2005. They were so impressed with our technology that they mentored and partnered our company. When the iPhone launched the world’s first 3G handset, we went in, and now we are about to see iPhone 12. This proves one thing – without technology, there is absolutely no opportunity. Research and development is the most important core factor for the survival of an enterprise – especially in the technology industry.
Huang: You had the opportunity to have a blueprint for future development at an early stage, which is priceless. Most of the companies in Taiwan involved in the chain of the international division of labour just make what other people need, following the others step by step, not being able to see beyond today’s landscape. Win Semiconductors, on the contrary, foresaw the future because it planned ahead of time.
Chen: Indeed, because of this advanced deployment experience, we still have a 10-year tech development blueprint in place. In addition, because of our position in the industry, all of our tier-one customers come to us to discuss the future, which means we have been talking to all of the top global players about what will happen in 10 years, so our blueprint is probably pretty accurate, and even today, our tech is prepared to serve them.
Huang: Are these technologies developed in-house or from external sources?
Chen: I made two bold and expensive decisions back in the day to take the long road ahead. The first one was technology. I insisted that we had to research and develop all the technologies ourselves. So far, we have developed almost all of our technologies, and we are developing all the III-V compound semiconductor technologies. So as we insist on our own research and development on the one hand, and diversified development on the other, it is very costly.
The second is production capacity. Technology alone without production capacity is like having a good kitchen without a stove and a pot. Therefore, our production capacity is always based on the demand and advance deployment, it takes us two years at the fastest from building the factory to shipping the products. We have to be very clear about the future, otherwise, the risk of expansion is very high.
Huang: In the global production chain, you are no longer an OEM [original equipment manufacturer], and your role is becoming more and more valuable, which means you are formulating specifications on the one hand and selecting technology on the other. By the time people behind you catch up, things are already set in stone. This is the key threshold for Taiwan’s future industry to move up to the next level, and you have to jump over it to make the leap.
Chen: The most popular words this year in Taiwan are “advance deployment”, to plan ahead, you must have the strength to do so, and the strength comes from your R&D investment. From the beginning, I have not set limits on investment in R&D, we are looking at the need, not at the budget – the latter is just a management tool.
I have strong views on key performance indicators (KPIs), too. KPIs can only be used in the executive unit, but as you get to the top of the strategic thinking, KPIs are of no use at all. KPIs are even more useless in the context of the R&D department.
Huang: In recent years, Win has made several reinvestments and mergers and acquisitions (M&A) – how are these transactions related to your core technologies and main products? You also have a corporate venturing subsidiary, Win Venture Capital, is it going to conduct financial or strategic startup investments?
Chen: I am not so sure whether it is financial or strategic, but I should say both, depending on the opportunity. If it is related to my work, such as SiC (silicon carbide) solutions for RF (radio frequency), then I will definitely invest in it, and this would be strategic.
I also pay attention to the biotechnology vertical – it will have its place in the future and will be a good opportunity for Taiwan. Our ancestors left a lot of wisdom in the field of Han herbology, but they did not use scientific methods to develop them. If we combine western knowledge or technology on this basis, there is actually quite an opportunity – if a startup is developing this, we will also invest in it.
In addition to drugs, there is also testing, tools and medical materials. Taiwan’s information and communications technology (ICT) industry is strong, and if we can combine it with biotech, ICT still has an opportunity in this area. So while I cannot say what the purpose is across the board, it will depend on the opportunities.
Huang: Would it possible for you to achieve some of your goals through M&A in the future?
Chen: Yes, I would not rule that out. In fact, we had merged with Global Communication Technology in the beginning – we were generating profit but they were not, and the share exchange ratio was their 1 share for our 1.5 shares – people thought I was mad for buying a company that was not yet generating revenues. However, I calculated that the difference was not hundreds of millions of [new Taiwanese] dollars, and I spend more than that a year just to fight off competitors, so this is a means to eliminate that rivalry. But if there is an opportunity that can expand our own strength, we will also consider that.
Huang: Finally, is there any advice you can give to other businesses from your own experience?
Chen: I do not dare to suggest anything because I only know my own industry. But I firmly believe all companies have their own virtues, and it is up to them to establish their core competency based on these assets, their own needs and the characteristics of their industry. Without a core competency, all other ideals are futile. Secondly, I would suggest that we should look further ahead, rather than just thinking about the current situation. If you focus too much on the current customers and only change things in this way, then you will be stuck and never get out – this is my humble opinion.
Founded in 1999, Win Semiconductors is the world’s largest GaAs foundry with 70% market share. The company employs about 3,000 people and has a total market capitalisation of over NT$120bn ($4.1bn). Its annual revenue in 2019 was NT$21.38bn ($729m).
Translated by Liwen-Edison Fu.