One Belt, One Road

China’s policy may have big implications for semiconductors and IP.

popularity

China’s so-called One Belt, One Road policy on global trade could have significant repercussions on semiconductors and IP if it succeeds. It’s hardly a slam-dunk, and China has been vying for a larger position in the semiconductor industry for some time. But this is a completely different strategy because China is not attempting to go it alone this time.

China has been rather quietly building an infrastructure over the past four years, constructing railroads that roughly re-trace the Silk Road for terrestrial trade, and establishing maritime routes to supplement it. The plan is to create an enormous middle class that spans 68 countries, based on the assumption that it will spur consumption of goods manufactured in China.

Trading blocks don’t always work as well as they do on paper, and political differences can get in the way of the best laid plans. Moreover, it’s becoming a lot more expensive to fund this effort than initially expected. When it was first introduced, China established a $40 billion Silk Road Fund, according to Mckinsey. The Economist reported in May that number has increased to $150 billion a year. That’s a lot of money for infrastructure, and at some point there has to be a tangible return on that investment.

Still, even a limited amount of success here could have significant implications for multiple industries if high-speed transportation and financial links become established. Semiconductors are a foundation for many of those industries..

Naveed Sherwani, SiFive’s new CEO, points to the opening of People Bank of China branches in Pakistan as a sign of China’s effort. “They will repeat this everywhere. They are even teaching Chinese in Pakistani schools. Think about what happens if all of these countries vote as a block in the United Nations. They will have a big impact on hardware and software. Whatever standard they adopt will permeate throughout these countries.”

SiFive is hoping this coalition adopts RISC-V as a standard. And for startups, using open-source hardware may make sense in some areas. But to be successful requires the development of a huge ecosystem of tools, IP, manufacturing, and continued investment to make this work. China’s previous forays into memory manufacturing and fabs didn’t work as planned, and some of these markets may require significant investments for extended periods of time before they show positive results.

There is value in an ecosystem. The key reason why ARM has been so successful against Intel is that it recognized the need for low power in mobile devices, and it created opportunities for partners to provide tooling, integration, software and support. The big question now is whether China can build a similar ecosystem with this coalition of countries.

Still, timing is critical, and it’s as much planning and good timing as it is luck. It’s much easier to displace an entrenched company when something is broken. ARM’s real breakthrough came when the center of gravity shifted from PCs to smartphones and tablets, something no one would have predicted at the turn of this century. Intel, which ruled the PC market for nearly two decades, was slow to embrace that shift, creating an opening that ARM continues to exploit. (Intel has since regained its footing in the data center and the cloud.)

Whether the IoT, IIoT, and new markets such as automotive, medical electronics and virtual/augmented reality provide a new entry point comparable to the shift from PCs to mobile phones remains to be seen. It’s not clear who will own those markets, or even where those battles will be fought. At this point, it isn’t even clear if carmakers will continue to dominate the automotive market once autonomous vehicles begin taking over. And where China fits into this picture is unknown.

But China isn’t the only one vying for a larger piece of the pie. Synopsys has its ARC processor line, Cadence has its Tensilica DSPs, both of which are targeting sensors and other new opportunities. Even Intel has recognized early on this time that there will be money in the various markets that collectively make up the IoT and IIoT. And Samsung is making a strong push into the most advanced nodes and advanced packaging, combining technology from multiple companies.

China’s One Belt, One Road is at the center of a large and so-far rather loosely constructed ecosystem. And after investing hundreds of billions of dollars, it’s still too early to tell how successful that will be. That may depend on when the money runs out. But it’s also hard to count out China. It has a history of playing the long game well. The original silk road was a conduit for trade for nearly 1,200 years, and China can trace its history back roughly 5,000 years. In contrast, the transistor has only been around since 1947, and the integrated circuit since 1960.