Comments by "shazmosushi" (@shazmosushi) on "Can TSMC Make Enough Chips?" video.

  1. Fortunately for TSMC and Samsung, the government of China's strategic investments are extremely wasteful. Unlimited money does not necessary lead to good outcomes (see SpaceX vs Blue Origin). The government of China has spent tens of billions over decades to build jet engines for passenger aeroplanes with little to show for it. They might eventually succeed, but due to US sanctions this is now unlikely. Consider that SMIC (China's TSMC competitor) can't even buy ASML machines due to being added to the Entity List. Huawei's HiSilicon can't license ARM Holdings chip reference designs. Huawei can't use Android due to sanctions. China is even being banned from accessing EDA (chip design software) from America. China could succeed in building indigenous replacements of these technologies one at a time, but when the United States is sanctioning everything at once it's VERY difficult. Especially when eg, Huawei's revenue streams starts drying up and those companies face big budget crunch. Xi Jinping overplayed his hand (around the usual issues: annexing the South China Sea, 1 million Uigher internment, crushing Hong Kong democracy, etc) and caused the US to react. Now if the US sanctions continue I think China will struggle to ever becoming competitive in the semi-conductor industry.
    16
  2. ​@King Lee I think China will eventually make 4nm chips, I expect it won't come soon enough to make a difference. But we can disagree though. Unlike constructing buildings, doing laboratory research, or even building space probes, large scale manufacturing of semi-conductors nodes using leading edge requires acquiring a lot of international equipment and expertise. And China is currently being blocked from the equipment (but not yet the expertise). I'll be very surprised if China is able to manufacture leading edge semiconductor nodes in the next 30 years. I should also mention that United States also does NOT have any leading edge semiconductor nodes.
    9
  3. The key indicator for an aging population is the 'total dependency ratio' (the working age population divided by the number of children and elderly) and the median age. The cost of an aging society is (at least initially) economic stagnation (while the rest of the world marches forward). Consider that Japan averaged 0% economic growth **1980-2020**. While high-tech industries like semi-conductors tend to require relatively small numbers of extremely highly skilled engineers, they work closely with a large numbers of skilled technicians and need extremely good communication. Even an aging Taiwan (and South Korea) can retain leadership in these industries if careful. The next 60 years are going to be brutal for Taiwan's demographics (and also for Korea, China, even the US). For more information search for the "United Nations World Population Prospects" website. Select Graphs and Profiles -> Taiwan -> Probablistic Projections -> Pop Ratios -> Total Dependency -> Age 0-19 plus 65 and over by 20-64. Well managed aging societies can retain their infrastructure and industries but once you're in a "low-fertility trap" with high debt it's difficult to leave.
    2
  4.  @etbadaboum  Immigration can definitely reduce the impact. I hope Taiwan (and Japan, South Korea) consider allowing more foreigners to immigrate and assimilate. But fixing the root cause is important: the cost of having children is too high for many families. Especially the cost of housing and child care. The developed world should at least achieve 2.1 children per family (the replacement rate).
    2