China vs. USA – Build the wall!

China doubles down on semiconductor unreality.

• While there is no doubt that China is in a better economic condition than most of its rivals its plan for self-reliance in semiconductors has very little chance of succeeding although its plans in AI look far more realistic.
• These plans were made official at the National People’s Congress last week where a GDP growth target for 2021 of 6%, a 7% increase in military spending and 7% increase in R&D spending for the next 5 years were all approved.
• The R&D spending increase is aimed at enabling China to gain technological independence from the USA (and other rivals) such that the sanctions which are destroying Huawei will become ineffective.
• RFM research (see here) has identified three areas of particular importance for China:
  • First, Semiconductors where RFM research indicates that China has very little if any chance of becoming self-sufficient.
  • The problem here is that China simply started too late by which time the key technologies had already been established.
  • In order to be truly self-sufficient, China has to replicate the entire supply chain of semiconductor capital equipment currently dominated by the USA, Japan and The Netherlands.
  • China can replicate TSMC’s fabs to its heart’s content but if it cannot make the equipment that needs to go in these fabs it will remain dependent on imports.
  • China’s efforts to date have borne very little fruit and have either been thwarted by government intervention or have already run into difficulties.
  • A good example of this is China’s attempts to replicate the Electronic Design Automation Tools industry currently dominated by 4 US companies.
  • These companies all have Chinese representative offices and China has been taking people from these offices and creating start-ups aimed at replicating the EDA tools.
  • Unfortunately, representative offices tend not to have the kind of staff that would be able to do this very quickly meaning that while China may eventually succeed in this endeavour, it will take so long as to be effectively useless.
  • Hence, RFM continues to think that China will remain dependent on imports for semiconductor equipment meaning that its strategy for self-sufficiency in semiconductors will continue to fail for the foreseeable future.
  • Second, AI: which is a sector where China is already a leader.
  • RFM’s research indicates that China is already well along the road to being a leading force in AI with 3 out of the top 5 companies in the world being of Chinese origin.
  • ByteDance in particular stands out as it has managed to create algorithms that are better at characterising and matching videos to users that will like them than anyone else out there.
  • This is why TikTok has been so successful both inside China as Douyin and overseas as TikTok.
  • This combined with a very strong position in autonomous driving puts China at least in an equal position when compared to its USA and European rivals.
  • Third, Quantum computing where China is also in a strong position (see here).
  • While the West is going down the superconductor route, China has chosen to focus on optical circuits where it has done a very good job of managing to scale the system to achieve colossal speed.
  • The net result is that China is claiming that its supercomputer is 380,518x faster than Google’s supercomputer which uses superconductors.
  • None of these systems can really do anything useful yet but this will change with time and once again, there is plenty of evidence to suggest that China will be a leader in this field.
• The net result is that the current battle is being fought in semiconductors where I think China has an insurmountable disadvantage, but this will not always be the case.
• Sooner or later, silicon will run out of road and new manufacturing techniques for computers will be required and here it is clear that China is comfortably in the running.
• It is in these new technologies where I see a wall being built where everyone is forced to pick a side when it comes to which standard to support.
• Networks that are isolated from one another result in much lower overall value creation than if they are part of a single larger network.
• For this reason, I think that the long-term growth in technology will be much lower as a result of this rivalry that shows no sign of abating.