Getting From Lab to Fab: a Chinese Challenge

Travelling the Silicon RoadSHANGHAI — One of the challenges peculiar to China as a developed nation but with a relatively nascent chip industry is that of getting technology out of the lab, into the fab, and into the hands of users.

And thereby making some healthy yuan in the process.

Even while government and industry officials, as well as executives acknowledge that homegrown IP is critical to China’s developing chip industry – not to mention its ability to meet its exploding domestic demand — it has had a poor track record in getting technology developed in the lab into commercial production. It is a task that is difficult at best – for every R&D achievement hailed as a breakthrough, few translate into commercial profit.

It’s a universal problem, noted James Gao, founder and president of Shanghai-based startup Apexone Microelectronics Inc., a fabless, high-end analog and mixed signal chipmaker.

“It not only exists in China,” Gao said. “It’s a similar kind of issue back in the States.” Gao studied in the United States and worked in the chip industry before returning to China to start Apexone in 2002.

The problem, as his company experiences it, is hiring fresh graduate students with great ideas — ideas they brought to fruition in the lab — who want to apply it to Apexone’s products. But the optimal power conditions used in the lab don’t exactly emulate the lower power demanded by a commercial mobile device – in other words, what worked in the lab often experiences problems or just won’t work under real world conditions.

“That is a major challenge for university lab students,” Gao said.

Narrowing the Gap

But it is apparently a particular problem for China. As some students and executives here suggest, part of it is the legacy of communism and state-owned business. It isn’t like China hasn’t had a chip industry for some time. But in the previous climate of a state-owned economy there was little incentive for innovation. And now, as China continues to convert to a market based economy, there is a gap between what Chinese chipmakers can produce and the leading edge in the rest of the world.

Most domestic chip designs originating in China today are at .35-micron to .25-micron, according to Hou Jinsong, manager of the software department at Beijing-based CEC Huada Electronic Design Co. Ltd. (HED). HED was the first fabless design house in China. Hou’s unit within the company is the first domestic Chinese EDA software developer.

Very few designs below 0.18-micron are produced in China. Below that parasitic problems are still an issue for most Chinese design houses, Hou noted. That’s why companies all along the food chain are courting foreign VC investment and joint ventures — to help fund the needed R&D in China.

“I think the gap will be narrower (in the future), but it depends on the investment,” Hou said. “If you don’t have enough (proprietary) IC designs, you won’t grow rapidly. For the EDA industry, investment is very important.”

Converting Brain Power to Economic Power

It’s not that China doesn’t have the EEs or the research programs at universities and institutes to create that necessary domestic intellectual property. It’s the conversion to the commercial market like Apexone’s Gao described that many here say is problematic, and perhaps a new idea for the Chinese, as they move from state-owned enterprises to a free-market system.

Take Beijing, for example. There is a reason Intel recently chose Beijing as the location for a new R&D center. The capitol of this country for centuries, it was cast and maintained as the capitol not because of abundant natural resources – which it doesn’t have – but because of its strategic geographic position.

Over the centuries and into the present it has become the heart and soul of Chinese art and culture and in many respects its brain. The central government is here, as well as many of China’s most prestigious and research institutes and universities, such as Tsinghua University and the China Academy of Science.

“The only difficult thing in Beijing is how to realize the potential of all the brains here,” said Liang Sheng, himself a possessor of a doctorate degree and the section chief of the Department of Information Industry within the municipal government. He is also deputy director of the Beijing Semiconductor Industry Association.

It is a problem that China has worked on for years. One that Liang and others in China suggest was one of the reasons the central government decided to move towards a market-based economy. While the situation has improved markedly in China in recent years, there is still plenty of room for improvement, Liang acknowledged.

But China is learning. One piece of evidence to support that is the start-ups that have been launched out of China’s universities. It’s not a lot, especially when compared to the United States or Europe. But it’s growing. The aforementioned Tsinghua University, for example, has spawned three large start-ups. Beijing University has launched two. North East University in Shenyang has launched six companies.

Many of those companies are retaining close ties with their respective universities, participating in and sponsoring related research programs. In some cases, the two are tightly conjoined, as in the NEU Information and Technology Co. Ltd., a spin-off of North East University (NEU), which also plays host to the Shenyang Embedded Technology and Engineering Research Center.

“The company is a platform for product development,” explained Deng Qingxu, a doctorate and a graduate of NEU, and VP of NEU Info. and Tech.

But while the situation has improved considerably in recent years, the problem of turning Chinese brains into Western-style profit still persists here in China, he acknowledged. “It’s not totally overcome,” he said, adding that probably no more than 50 percent of what gets developed in research labs here makes into a commercial application.

He observed that government support still tends to focus on pure research, as opposed to letting market demands drive research. And the government is still aggressively pursuing research, to be sure.

In addition to encouraging innovation and IP development through the tried and true method of market demand, it is putting a lot of financial resources into R&D. On one hand, it is doling out money to support companies large and small with promising IP. Semiconductor Manufacturing International Corp. (SMIC) is an oft-mentioned example of this. SMIC is manufacturing 90-nanometer designs for certain customers at its 300mm fab in Beijing, the only 300mm fab in China.

Another method is through government grants. The central government has set aside some $16 million (130 million yuan) a year specifically for system on chip (SoC) development, for example. The government is also currently building two R&D centers – one in Beijing and one in Shanghai.

Electronic News Travels to China“In my opinion, research should be more practical; there should be more cooperation with industry,” Deng said. “That way you find out what the urgent needs are and concentrate on them.

“Just following random ideas in white papers, this isn’t fruitful,” he added, suggesting that relationships between research centers and industry could be tighter still. “As a researcher, you should take your cues from the factory.”

Editor’s Note: As explained at length elsewhere on this site, this is a news story written by me that originally appeared on the now-defunct Electronic News’ website, which is long gone. It’s former sister pub Electronic Design News (EDN) currently holds the copyright to all Electronic News copy (to the best of my knowledge). You can still see a copy of this story at EDN.

Leave a Reply

Your email address will not be published. Required fields are marked *