China wants global semiconductor leverage
Posted:
05 Sep 2014
In the past weeks, two news items involving
Intel and
China stirred up the investment industry, as well as attracted Chinese media attention. First is that Intel is eyeing China's Spreadtrum as an investment to shore up its mobile business, according to a Chinese news site. Second, Intel should consider acquiring Taiwan's MediaTek as urged by RBC Capital Markets analyst Doug Freedman.
Reached by
EE Times on Tuesday, Intel spokesman Chuck Mulloy was, of course, mum, other than noting: "We don't comment on speculation nor do we speculate on what might have caused the speculation."
This is a predictable no-comment. Both items might indeed be pure "speculation" worthy of no further interest.
However, it was only a few weeks ago when a group of Chinese investors—including a state-owned firm—offered a buyout proposal to US digital imaging chipmaker OmniVision. Putting together a string of recent events, we couldn't help but wonder:
- Does selling Spreadtrum to Intel, or allowing the foreign chip giant to invest in the now state-owned company, make sense for China?
- What's the real motivation behind China's wanting to acquire OmniVision?
- Do any of these moves have anything to do with China's recently unveiled "National Framework for Development of the Integrated Circuit Industry"? (After all, the Chinese government is setting up a huge annual investment fund to support the nation's semiconductor industry.)
Some US-based electronics industry executives told us that a Spreadtrum sale and an Omnivision purchase are both among the potential financial plays contemplated by China's investment community. Others, however, are sceptical of the whole scenario.
EE Times has been scrambling to connect the dots of China's seemingly unrelated investment moves in recent months.
Nicky Lu, chairman of Etron Technology Inc. in Hsinchu, Taiwan, is one of those industry executives convinced that these manoeuvres are closely tied to China's national IC industry framework. He says they make perfect sense.
Just to be clear, we regard Lu as "the man in the know."
While serving as chair of the Taiwan Semiconductor Industry Association (its politically correct name is "Semiconductor Industry Association in Chinese Taipei"), Lu earlier this year became the chair of the World Semiconductor Council (WSC).
Lu explains that China's new policy, different from those in the past, is the infusion of private investment funds. It allows professional financial investors to bet on which entities—fabless, foundries, and/or research institutes—deserve the funding.
In Lu's view, if the Chinese funds actually succeed in improving the value of Spreadtrum and manage to sell it off (to Intel or not), China wins.
China's proposal to buy OmniVision, on the other hand, will have further impact.
By taking over the world's leading CMOS image sensor vendor, China will gain instant access to the global market and the company's formidable market share. More importantly, such a deal generates demand for volume production of CMOS image sensors in China (not in Taiwan)—enough to fill the capacity of home-grown Chinese fabs like Semiconductor Manufacturing International Corp. (SMIC) based in Shanghai.
In short, China's investment funds are on the lookout for acquiring successful companies in the global market.
How China funds are allocated. Source: Data compiled by EE Times based on media reports in China and Taiwan, and interviews with industry sources
As illustrated in the table above, as much as $98 billion, to promote M&A activity, will flow to local governments and their regional private equity investments in China. This is in addition to government funds for national IC industry support.
Modelled after the Beijing IC Industry Equity Investment Fund, Chinese provinces including Wuhan, Shanghai, and Shenzhen are racing to build "regional" private equity funds. Rivalries between different regional private equity funds were already evident when the Beijing-based Tsinghua Unigroup outmaneuvered the Shanghai Pudong Science and Technology Investment Co. (PDSTI) to buy Shanghai-based Spreadtrum. Meanwhile PDSTI, not to be outdone by Beijing, unveiled a plan to acquire Montage Technology of Shanghai. PDSTI is also a part of the investment group that has offered to buy OmniVision.
China's national blueprint for semiconductor industry development is far from the stodgy, top-down model of the planned economy era, according to China hands familiar with the industry. Today, it's much more "market-driven."
A case in point is that the money made available by the government and China's private investment funds isn't just for investing in Chinese companies. The funds can go global to acquire technologies and companies with the best potential to expand China's semiconductor industry. The OmniVision deal would fit that bill, according to Lu.
China to follow Taiwan's playbook
It's well known that the Chinese government isn't happy about the widening gap between the number of chips China imports from multinationals and the volume China produces on its own.
China's integrated circuit consumption and production comparison. Source: "Continuing to Grow, China's Impact on the Semiconductor Industry, 2013 update," PwC
If history is any indication, every nation has harbored similar concerns. In other countries, the solution has been a national industrial policy to develop home-grown semiconductor production.
Lu observes that China has decided to follow Taiwan's playbook, rather than pursuing models practised by Korea, Japan, Europe, or the United States.
The Taiwan playbook, in a nutshell, is about creating a service model, says Lu. Taiwan first launched the foundry business, including such companies as Taiwan Semiconductor Manufacturing Co. (TSMC) and UMC, as a mother ship. Taiwan, in parallel, developed additional service-related businesses ranging from design services to packaging, equipment maintenance, and wafer-level testing companies. The mother ship, then, spawned a cluster of IC design houses—described by Lu as "a fleet."
It took Taiwan 15 years before consumption and production of chips got even, says Lu. Back in 1990, Taiwan's National Subµm Project, led by Taiwan's Ministry of Economic Affairs, began. The project enabled Taiwan to build 8-inch wafer technology for design and manufacturing, thus facilitating growing investment in the semiconductor industry in Taiwan. In 2005, the semiconductor exports exceeded the country's total imports.
Source: Etron Technology
Lu predicts that it will only take 10 years for China to reach that break-even point. The next couple of years, however, will be critical, he feels. China's policymakers hope to show early signs of growth in a key business, in employment and technology, that matters to China's economy.
Who's behind the national plan?
As Lu tells the story, those who have drawn up the National Framework for Development of the Integrated Circuit Industry are an elite team—fewer than 10 people. Consisting of physicists, private equity people, and technology experts, the team composes the national semiconductor office, which sits right next to China's prime minister's office. Their goal is to find the most effective and meaningful path for China's semiconductor industry to plot its own future.
"I have a lot of respect for the intelligence of policymakers in China," says Lu.
Of course, a national plan without engineering talent can only produce empty promises. China, however, is ready to up its game, according to Lu, both in terms of the fundamental science and the volume of talented engineers to carry out some of the toughest projects.
He cautioned us: "Don't judge China just by looking at today's China." More important is what China is capable of doing in the next 10 years.
Consider the example of the fundamental science for developing a 3D topological insulator, says Lu. A project that originated at Stanford University now has a group of scientists at Tsinghua University (Prof S.C. Chang's group) and the China Academy of Science vigorously working on it.
Further, take note of the 400,000 students graduating from universities in China every year with engineering degrees. Fifty per cent have degrees in integrated circuits, according to Lu. Where else in the world could higher education pump out so many students with engineering degrees?
Accountability
Let's face it. China doesn't exactly have the best record of accountability when government gets involved in spreading around investment money. Even projects that started out with the best of intentions have failed as money disappeared somewhere along the food chain. How do you ensure the accountability of the National Framework for Development of the Integrated Circuit Industry?
Several things need to happen, notes Lu. First, policymakers must show results. They need to demonstrate that the policy can generate better domestic jobs. They also need to encourage China's system companies to design with parts produced in China. The use of domestic components is especially encouraged for "special-use" products such as military applications. Policymakers also need to demonstrate the use of the domestic chips in such applications as IoT and healthcare systems that will be particularly helpful in improving the lives of Chinese people, he says.
Source: PWC
By choosing credible and successful companies (foreign and domestic) as targets for investment, China is seeking to produce results quickly, Lu says. That's why he views the next two years as critical.
Unlike the old China, which often insisted on developing domestic versions of global standards (e.g., VCD vs DVD) so that it could avoid paying licensing fees or royalties, the success of China's new policy hinges upon following international rules, Lu points out. Instead of pursuing a domestic standard for the sake of, say, a home-grown operating system, China must put commercial deals first.
In sum, China is at a threshold. China is ready. If it succeeds in the OmniVision acquisition, China can easily take OminVision's business away from TSMC and bring it to SMIC. As far as Lu is concerned, "The game is fair, and things in China are moving faster than ever."
Other regions in the world, on the other hand, have only themselves to blame if China bypasses them. As Lu concluded, innovation is the only way for us to differentiate ourselves.
In 2013, China's semiconductor consumption market grew by 10.1 per cent (more than double the worldwide market growth of 4.8 per cent). It has reached a new record of 55.6 per cent of the global market. Ongoing global demand for smartphones and tablets is the main reason for this continued strong growth in China's semiconductor consumption and will continue to be a factor in the coming years.
China wants global semiconductor leverage