• Skip to main content
  • Skip to search
  • Skip to footer
Cadence Home
  • This search text may be transcribed, used, stored, or accessed by our third-party service providers per our Cookie Policy and Privacy Policy.

  1. Blogs
  2. Breakfast Bytes
  3. SEMICON China Panel: Faster Than You Think
Paul McLellan
Paul McLellan

Community Member

Blog Activity
Options
  • Subscribe by email
  • More
  • Cancel
China
semicon
semicon china
semicon west
Breakfast Bytes

SEMICON China Panel: Faster Than You Think

22 Aug 2017 • 7 minute read

 breakfast bytes logoThe afternoon of the first day of SEMICON West was taken up by a series of presentations and a panel session called the China Innovation and Investment Forum. The participants were:

  • Lung Chu, SEMI China (moderator)
  • Hing Wong of Walden International
  • Tammy Kiely of Goldman Sachs
  • Allen Wu of Arm China
  • Huang Weixu of Hua Capital

Lung Chu, SEMI China

China is the largest and growing share of the worldwide IC market, growing faster than the world GDP. This year, for the first time, IC design passed packaging and test, but manufacturing is still relatively low. Having said that, there are 24 fabs under construction in China so the manufacturing numbers will change dramatically. Currently, the gap between local demand and local supply is widening, not closing, so the trade gap is widening. Presumably as all those fabs come online then that will start to change. 

I already covered some of the other points that Lung made in my overview SEMICON: China, China, China, so I won't rehash them here.

Hing Wong, Walden International

Hing Wong is the Managing Director of Walden International. Of course, Lip-Bu Tan, Cadence's CEO, is the chairman. They are one of the biggest investors in China. Hing himself has been at Walden since 2005, following 15 years in the IC industry. Walden is in its 30th year of operation, with their 100th IPO coming up soon. Walden's current fund has $300M for Silicon Valley and $300M for China. They also set up a larger fund for M&A. Roughly half their investments go into semiconductor and its ecosystem. 

Hing sees this as the era of AI, with cloud on top of what he calls "platform devices." Ther are three major platforms: personal devices, home AI platforms, and cars. All of these are semiconductor + AI.

Automotive was not an era for VC investment, but now it is. Apple is rumored to be getting into it. Even a relatively small player like BMW is as big as iPhone. In China, there are 1000 smart car companies, all electric, and all driven by IT people, not automotive people. China is a big market with over 50% of the cars sold there also developed there.

 Overall, there is a big shift in the center of gravity for semiconductors. Only one region is rising, China, and everyone else is declining in market share. That's why Walden invests in semiconductor in China. Today, only 5% of semiconductors are designed in China but it is growing at 25% per year. It is moving from a cost play, to a performance/feature plan. This is driven by the fact that there are now lots of engineers in China with 10 years of experience, which wasn't at all the case a decade ago.

Society in China is changing faster than ever before:

It will happen faster than you think

Tammy Kiely, Goldman Sachs

Tammy is the Global Head of Semiconductor Investment Banking at Goldman Sachs. She is based in San Francisco. She talked about The China Opportunity: A Banker's Perspective.

Tammy KielyShe started off looking at how other countries entered into the semi industry. Japan got DRAM leadership in 1987. Korea came in the same way, getting DRAM leadership in 1992. Almost none of this was driven by domestic consumption. For example, only 7.5% of Samsung's revenue used to be from Korea.

Today many purchasing decisions are made outside China (as an obvious example, Apple in the US decides which chips go in the iPhone, even though that represents a huge dollar value of Chinese imports of semiconductors). But China has lot of capital. It has another advantage too, which is that PE ratios are high (49X to 20X versus much 13X to 15X in US). A big proportion of capex for equipment is happening in China (those 24 fabs all need to be kitted out with tools).

The foreign versus domestic ownership is moving, too. In 2016, only one-third of 300mm capacity was Chinese owned (a lot at SMIC) but now there is a lot more domestic investment.

Capital controls impact China outbound M&A, as does regulatory concern in Washington. China has cut back on outbound investment, by 40+% between Q1 2016 and Q1 2017, which is a big cut. In the US, it is important to keep and eye on CFIUS (if you don't know what that is, see my post The Four Ts: Trade, Tax, Talent, and Technology Funding) where a lot more filings are expected and, right now, it is backed up with deal review.

Huang Weixu, Hua Capital

HUA capital what we learntHuang also talked about M&A, and the implications of the high domestic P/E numbers. He started looking at some of the big recent acquisitions. China buyers, with their overvalued/highly-valued (depending on your point of view) stock has deeper pockets. But when it comes to buying overseas companies, especially in US, there are implications from both capital controls and also CFIUS review risk.

For a US company to sell a business unit into China, the best structure is a joint-venture (JV) which gives access to the higher valued Chinese stock market. The US company gets the advantage of a higher valuation/exit for the BU and establishes themselves as a local player in China, with closer ties to the ecosystem. The steps to do this are to sell part of a BU to private equity (hey, guess what, Hua Capital is in PE!) and then convert the rest of the BU to shares of a publicly listed company in China and become a strategic investor. The PE then exits. In the case of manufacturing, a lot of the money actually comes back to the US (and rest of world) since semiconductor equipment is all manufactured there.

Panel Session

I won't try and cover everything that was discussed in the panel session. I'll point out some key points that I thought especially significant. I didn't always note down who made which points so I'll just leave them all anonymous.

A Chinese strength is a very disciplined education system. This is a good match for manufacturing, but might not be so good for innovation. US companies cooperating with Chinese have complementary skills, even before looking at cultural differences.

US companies face a big challenge in the next 5 to 10 years. It is no longer all about "selling stuff" and just putting more sales offices in China. There will need to be more JVs. Chinese management is now on a par with multinationals so it is no longer necessary (or effective) to dump an expat team into China. Chinese companies used to lose experienced managers to multinationals all the time and now that rarely happens. It is all Chinese companies when people move on. Things are moving fast so that going forward, the China business of US multinationals will be competing with local Chinese companies. Current management structures won't survive if US companies continue to manage the way that they have done historically.

China rising in 2010s is nothing different than Japan in the '60s and '70s, or Korea in the '80s and '90s. For semi equipment business, companies like Lam Research have 70% of revenue coming from Asia. The world is so globalized, especially in semiconductor, that it doesn't make sense to protect each country's boundary. If you go to any technical conference in the US, more than half the people there were born outside the US. What is happening right now is a knee-jerk reaction in the political space, which just highlights that Washington is too far from Silicon Valley to understand it. Talent is now flowing back to China to help build the industry in the same way has happened in Taiwan 20 years ago.

General Motors' JV in China sells more cars than in the US...and the margins are three times as much. It is an opportunity, not a zero-sum game. The valuations and exits are also an opportunity. To show how crazy it can be, Mediatek invested in a fingerprint company and at one point the valuation was higher than Mediatek itself, which is a big disparity.

Data and deep learning is another area where China may have a cultural advantage. Chinese are not really privacy oriented, with 55% of Chinese quite willing to share data, whereas in the US (and especially Europe) it is half that. So that could drive AI hard.

SEMICON China

Lung wrapped up by pointing out that SEMICON China is March 14-16, 2018, in Shanghai. It is the largest gathering of the semiconductor industry in the world, with over 70,000 attendees

Sign up to get the weekly Breakfast Bytes email: