A microprocessor war on Taiwan?

chiny24.com 3 years ago

Comments and forecasts regarding the imminent ("almost certain" – as 1 of the Polish authors wrote) China's invasion of Taiwan to take over microprocessor factories, scope further limits of hysteria and absurdity.

In the U.S. Army War College Quarterly quarterly, an article even appeared “Broken Nest: Deterring China from Invading Taiwan” which recommends Taiwan's “burnt earth” tactics, that is, complete demolition of TSMC factories in the event of invasion.

Well, I have 2 news for you: good and bad.

The good news is there will be no Chinese invasion of Taiwan due to problems on the microprocessor market.

The bad news is that the problems in this marketplace will last 2 or 3 years, then it will get a small better in the short term, and then it will get worse again. In the average word (10-15 years), supply problems will end on the market, but prices will rise. Then (15+ years) prices will fall again.

Why won't microprocessors origin the Taiwan War? The above-mentioned article does not take account of 2 fundamental points.

One of them can inactive be justified in my opinion, due to the fact that it is ideological, but the another is not. I'll start with the another one.

The microprocessor marketplace is much more complex than many seem. The most crucial thing for this analysis is to realize that

first: there are various types of microprocessors;
Secondly, their production is an highly complex process requiring respective 100 stages of usage of machines, software and semi-finished products, whose manufacturers are highly distributed (locations in very distant parts of the world), and more frequently they do not have competition (monopoly has grown on the basis of their technologies).

China is presently able to produce microprocessors in its own technology up to 14 nm. They work on up to 7 nm technology, have no chance of producing 5, 4 or even 3 or 2 nm technologies.

Nevertheless, it is worth noting that microprocessors up to 14nm are about 80% of the market. They are sufficient, for example, for 5G applications (base stations and another ground equipment), NEV vehicles (New Energy Vehicle), industrial robots, appliances and many, many others. The problem is 5G in smartphones and applications requiring 7nm or 5nm technology – tablets, laptops, etc. Of course, the fact that China has factories capable of producing up to 14nm in technology does not mean that they are able to full meet request on the Chinese market. That's far from it.

A separate issue is the supply chain. As I mentioned, it is highly monopolized and dispersed throughout the world. We have here de facto monopolies

for Lithographic Machines EUV (Dutch ASML),
two companies producing microprocessor plan software (both American),
two companies producing chemical reagents for etching waffles (both Japanese),
one or 2 companies specialising in “growing” the waffle,
And so on, and so on.

The strong mastery of 1 or 2 factories (the concept of China's aggression into Taiwan) does not solve any problem. So what if we have machines (I skip the question of possible sabotage and evacuation of highly qualified personnel operating machines, whose operation is know-how and many years of experience) and software, if there is simply a good chance that replacement parts will run out of them in a moment, will degrade the software, and we will run out of intermediates for production? After all, it is highly likely that after China's invasion of Taiwan, ASML will halt the supply of spare parts and service services and Japan will halt the supply of chemicals.

Let's get back to the first, ideological issue. In my opinion, this is due to the projections of typical US policy patterns and their usage as a cripple in assessing the motives and calculations of the Chinese authorities. The framework of this article does not let a long-term analysis of this subject, so I will just say that China resorts to strength solutions as a last resort, has a prepared and thoughtful plan “what then” and mostly considers the usage of force to be the worst of all available means. It is clear from these calculations that the acquisition of TSMC's factories is not going to do anything, and secondly, the issue of "thinner" processors than 14 nm only concerns 20% of the market, and thirdly, the costs of military operations, business of Taiwan, economical sanctions, etc. are so advanced that it is better to spend this money on building complete logistics chains for the production of microprocessors. And that's what China chose. That's what they do.

I do not have a place to describe in item the Chinese activities related to the construction of its own full independent and complementary ecosystem of the creation and production of microprocessors – we request to give a fewer more words about this bad news mentioned at the outset – but I will just say that there is no week that there is no information in the Chinese media about further investments, companies, peculiar industrial zones, purchases, joint-venture, etc. on 1 of the elements of the supply chain needed for the production of microprocessors. 2 years ago, a peculiar government squad was set up to identify all stages and elements of specified a supply chain and to identify all the "narrow throats" that China sees in it. Subsequently, they were systematically taken over for their liquidation. fresh factories are being developed, own EUV lithography machines and another production stages are being developed, fresh chemical plants are being created for the production of chemical components, work is underway to make their own software – the list is long. At the same time, Chinese scientists search alternate solutions – photon microprocessors, based on gallium nitride, graphene, nanomagnetic switches, etc.

What does this all mean for the world's microprocessor markets? In my opinion, the most crucial origin in the short-term is not China's invasion into Taiwan, but the US policy. It is about based on attempts to slow down the technological improvement of China by extending the ban on the sale of microprocessors (to be added to the ‘black list’) to another Chinese companies in subsequent industries. This is not only about 5G, but besides about drones, cameras, artificial intelligence, and in a minute it may be about NEV vehicles and solutions related to digital yuan. The effect is that almost all Chinese companies that request microprocessors in their products (and Chinese production presently caters for between 16% and 20% of their needs), build warehouse stocks (from 6 months to 2 years!). For they do not know the day or hr erstwhile they can find specified a "black list". Let us add to this the disruption in the supply chains caused by COVID and the time needed to build fresh factories (counted in years) to meet specified increased request (remember about 80% of the world's request for microprocessors is China) and we have a ready-made recipe for current short-term problems.

Then fresh factories will be built, China will accomplish independency in the production of microprocessors (at each stage) – the ASML president estimates that they request about 15 years for this. The marketplace situation will stabilise somewhat. However, the US policy of slowing down China will lead China to make its own independent technologies (from physical production, through a set of device instructions to its own microprocessor designs) and diversify the microprocessor market.

Awfully simplistic – the West will have its standards and factories (mainly in the US and possibly respective in Europe), and China will have its own. China will produce for the interior marketplace + Africa + any countries of the RCEP agreement (ASEAN countries) + another countries for which the price will be more crucial than ideology. Therefore, there will be a simplification in R & D expenditure in this manufacture in the West (smaller marketplace = little money for research), a simplification in economies of scale (less production), and thus higher prices for the final product. This will, of course, translate into the cost of microprocessor devices (i.e. almost everything “technological”). These "Made in China" will be comparatively cheaper due to shorter supply chains, a more advanced industrial revolution 4.0 and a more thoughtful and pragmatic state policy on "microprocessor policy".

Editorial: Leszek Slazyk