An EDA Veteran Endeavors to Take Design Services to the Next Level

DigiTimes - December 1, 2004

Jim Su is an EDA veteran. He worked for Cadence Design Systems, based in the U.S., for 10 years in the 80s before being assigned to be the Technical Director of Cadence Taiwan. His four years there coincided with a growth spurt in Taiwan's IC design industry, as well as a 10-fold increase in Cadence Taiwan's revenues. In 1998, Su decided to leave his long Cadence tenure behind and start his own business. However, right after he founded EE Solutions in 1999, he had to face an unprecedented slowdown in the semiconductor industry and price wars with many other local design service companies. With 20 years' experience under his belt, Su constantly pondered how to make a breakthrough and expand EE Solutions during his frequent drives between Taipei and Hsinchu, developing his own vision of the future of the design service industry.

Design Services Survival Skills

“IP, fab capacity, and design methodology are the three elements of survival in the design services industry.However, services are hard to reuse, and it's also hard to control resource allocation. These are the challenges that the design service industry must face.” Su claimed, insightfully pointing out critical areas facing design services operation.

According to Su, the design service industry began when Cadence Design Systems established its Consultant Services Division (CDS) in the early 90's with the goal of providing chip designs to customers. Starting in 1995, Cadence merged with 19 design factories one after another in order to diversify its design service business. Accordingly, the revenue from the service also grew from US$50 million to US$400 million by 1998, a sizable part of Cadence' total revenue of US$1.2 billion for that year. Cadence later spun out their design services division, forming a company named Tality, which employed 4,500 people at its highest.

Recalling the industry history that he personally experienced, Su said: “When a design service company reaches such a scale, many problems emerge. ‘Service’ is not the same as a ‘normal’ product, as we can only charge once for completing customer's designs, unlike semiconductor companies that can repeatedly use the same design when manufacturing chips to increase revenue.

“We also can't control when and how we get orders.” Su joked. “Customers don't place orders at your convenience.”The labor cost and the inefficient resource management eventually led to Tality's collapse, with no choice but to downsize and eventually be reabsorbed into Cadence.

The experience made him realize the importance of making the service industry ‘reusable.’ “Only through establishing a wide range of IP portfolio will this be possible.” said Su.

In addition, if one has a stable wafer capacity, especially when coupled with proprietary IP, designing a turnkey solution for customer manufacturing is the best way to establish a close partnership with the customer while maximizing reusability. However, with countries like China and Malaysia entering the semiconductor manufacturing industry, Su believes that the importance of holding wafer capacity will decrease in time.

Su thinks that, as a whole, the design service industry exists due to the lower cost of outsourcing to design service teams, more efficient resource use for customers, which results in better designs. “IC companies should focus on product specs and market planning and let us do the design work.” Su said. “Each and every step of the design process is important. Wwe must have a solid methodological foundation to build a strong design team.

“If you look at it from the opportunity standpoint, entering the deep sub-micro era, the trend to use design services is obvious, as typical companies can't afford to design chips on their own anymore, except the biggest companies. Just think about it; the smaller IC designers can't afford to buy expensive EDA tools continuously.”

Getting Farther Ahead

Although more complex design requirements create growth opportunities for the design services industry, many IC designers fail before reaching this more advanced technical stage. Su once heard it said that, of Taiwan's 300 IC design companies, because of their small scale don't make a profit, 1/6 are struggling and only 1/12 make money.

Su said: “If customers can't make money, how can design services providers make money? It's been very difficult for Taiwan's design service providers, especially because there are more designers than the market needs. Price competition is common. To avoid getting trapped in a price war, we must target different customer groups. Only when we find the right customer base can we achieve win-win solutions.”

Therefore, Su tried to target the U.S. and China markets. “I had some U.S. customers. Although I didn't make much money from them, I learned new techniques.” he explained. “EE Solutions developed 10-million-gate NPU products for them, so the current 3 to 4-million-gate WCDMA base station chips are not difficult at all for us. It was important and valuable for us to invest, to explore new markets and to accumulate technology in the U.S. In order to fight a war, you need to first have solid technical foundation.”

In the China market, EE Solutions' biggest customers are all local companies, especially big communications system houses. Su said: “It is especially difficult to forge close relationships with local China customers. We spent a lot of time strategizing; through our many field experiences, we gained the customers' trust. Now we have solid business.” In addition, due to China's aggressive development of their 3G technologies and markets, Su is very confident that EE Solutions can forge ahead alongside its customers, thanks to the accumulated technology and business channels.

Many surveys show that with costs increasing, more and more OEM manufacturers are now adopting standardized products (ASSP) instead of developing their own ASIC. When asked about this trend in the design services industry, Su thought that, as a whole, due to the shifting of design and manufacturing towards Asia, while U.S. design services may be impacted, Asia’s ASIC needs would increase, not decrease.

From a strategic point of view, Su emphasized the current ‘system houses’ are his primary targeted customers. For example, he would work with system houses to develop and replace existing chips. “Our goal is to help system-house customers become more competitive and specialized.”

Su observed that many system houses in China, especially communications companies, are trying to break into the IC arena by defining their own specs. He foresees that the big system houses will fund new IC design companies through investment or spinning off their chip businesses. “Taiwan's system houses, such as BenQ, Acer and D-Link, all followed this strategy. Forming alliances with these spin-offs in the initial stages will be the key business opportunities for design service providers.”

Therefore, Su will focus on large-volume business opportunities that can replace European and American standard parts. “By 2008, we'll be able to achieve this goal with many parts used in such hot products as digital televisions and cellular phones. The market is actually very large, it all depends on whether you can grab a slice of pie.”

“Simply put, we have to use functionality and reliability to assist our system house customers.” emphasized Su. “And at the same time, team up with them to beat IDM players.”

After several years of hard work, EE Solutions has reached NT$400 million in revenue. EE Solutions was founded on its own investments, unlike many other startups. Not until 2003 did WK Technology Fund (a leading VC firm in Taiwan with US$600M fund under management)and UMC Capital invest in the company. “Highly reputable investors will help to expand business.” said Su.

When asked why he came back to Taiwan, Su said it's only natural for ‘fallen leaves to go back to the roots.’ During his over 10 years at Cadence, Su experienced how Cadence became the EDA industry leader through continuous mergers and flexible business practices. Su plans to use the same tactics, pushing himself to enter bigger markets. Looking towards the future, Su humbly said: “Making a living isn't a problem.” But he believes that design services will eventually merge, so Su has to constantly search for ways to run EE Solutions as a long-lasting business and to create new business for the design services industry.

“At the moment, it seems the best model is to target system house business, strengthen customer collaboration, form strategic alliances and create mutual success. However, the industry is always changing; we must constantly inspect and review the feasibility of our strategies and modify them accordingly. As long as we try our best, even if we fail, we will know we gave it our all.” he said lightheartedly.

Jim Su

President and CEO of EE Solutions
Technical Director of Cadence Taiwan
Senior Manager of Cadence USA
MSEE, University of Santa Clara
BS in Transportation Engineering, National Chao Tung University