A few weeks back we wrote about RF Semiconductors. We closed with a section on China RF companies:

As with all things in technology today, we also have to consider what is happening with RF in China. There is obviously a lot going on in semis in China today. China has set out to reduce its reliance in foreign semiconductors, and this includes RF

Since then we have had a few people ask us about Chinese RF companies, and we wanted to expand a bit here.

Quick recap. Chips for Radio Frequency (RF) are the messy corner of mobile phone electronics. They straddle the divide between analog (i.e. Real World) signals and pure digital computation. We say messy because these tiny modules can have dozens of individual components buried inside them. Mapping out market share in this industry is a lot like trying to explain a map of the Holy Roman Empire in the 17th Century. There are all kinds of overlapping boundaries, with each domain operating under its own particular rules. Just as in that time, RF markets each seem to be governed by a different set of practices.

Put simply, there are two types of RF products – active and passive. Active parts are chips as we usually think of them, they do some form of processing or decision making. Passive parts are what you get in a basic electronics kit – resistors, capacitors, diodes, etc. These respond to electrical signals in a deterministic fashion, with no ‘logic’. The difference for RF is that passive parts use incredibly sophisticated manufacturing techniques and have to operate under very narrow, demanding specifications.

For reasons we have discussed elsewhere, the most strategic RF components today are passive filters. There are a limited number of companies that can build filters, owing to a combination of demanding manufacturing requirements, economies of scale and IP licenses going back twenty years.

When we said there were several Chinese companies building RF chips, we were referring to companies like Lansus, Vanchip and Unisoc. These companies all build active components, notably Power Amplifiers (PA). There are several others, including Huawei’s HiSilicon unit as well.

One the one hand, this is an important change to the market. The market for PAs has been dominated by companies like Skyworks and Qorvo of the US for over a decade. So this is one area where China’s push for domestic semis companies has succeeded.

That being said, as far as we can tell, none of these companies make filters. So they key high-value components still need to be imported from foreign companies. We strongly suspect that there are a few Chinese filter companies, but they do not seem to have made many inroads into smartphones, at least not at an international level.

What we find intriguing about these companies is how much the Chinese RF industry resembles the global RF industry 10 or 15 years ago. Back in the days of 2G, there were 20 or 30 companies making PAs. Every transition of wireless standards (from 2G to 3G and 3G to 4G) forced waves of consolidation to the point that today there are only five vendors left outside China. By comparison, China probably has a dozen PA companies, many so small that they likely exist on a custom design basis (i.e. not scalable businesses).

Ultimately, the industry can only support so many vendors, especially with smartphone growth now slowing to GDP levels. We think this dynamic repeats across all of China’s 1,300 semis companies. Someday, they will consolidate down to a more manageable number. The question is how many of those companies will grow large enough to become global players.

Photo credits:

Holy Roman Empire Map: WikMedia

Electronics Kit – Mizuei by way of Amazon