They can just allow the rmb to appreciate 5% per year for the next decade. This would send US inflation way up and their stock market into disarray. The large amount of money printing will tank the $US.
*Yawn* tell me something you haven't said in the last 10 years
Even if TSMC could do all that to maintain chip production, how could they ship their products to customers overseas, breaking a blockade that mainland China could impose?
This is one of those examples that pop the Psy-propaganda often making runs in US centered defence circles that insists US is in possession of some advanced classified technologies.
There's always risks of contamination when doping but it's a doable task. This was already accomplished by the research team I referenced earlier who created vertical tunneling transistors. The research proved that graphene could be used for logic albeit with a lot of jerry rigging. That doesn't mean graphene is a realistic silicon alternative. That's why other 2D materials that are actual semiconductors are researched in parallel and touted as having actual potential like platinum sulfide. I would use the analogy of nuclear fusion research which is pie in the sky research while thorium salt reactors are becoming a proven commodity soon to be commercialized.
I have a theory that IF China can commercialize quad patterned 7nm DUV process, that it will collapse margins for every process node below 7nm including the leading edge. Sometime after 2026, the excess semiconductor capacity that will be laying idle combined with what will probably become a largely self sufficient China semiconductor market above >=7nm, along with China's supply chain cost advantages, will totally destroy the global semiconductor business model. The price/performance curve between what will become China's DUVi 7nm chips and all leading edge EUV performance chips, will not warrant the price discrepancy that currently exists between the leading edge processes and processes 3 generations older. Imagine having chips performing at say ~50% of the leading edge, consuming say 200-250% the energy, but costing 5-20 times less.
It won't matter that China isn't at the leading edge when this happens. What will happen is, those at the leading edge will either be selling way less than they planned thus making way less money OR they will be dropping their prices to price discovery, collapsing their margins in the process and making way less money. Yes, there will always be sales of the very best, most power stingy, but most sales aren't in this category when the alternative will be literally 5-20 times less expensive. Either way, they will be earning far less income than what they expected with planning their capital and R&D budgets. I think mergers and bankruptcy are on the horizon for many of these leading edge semiconductor companies.
There's always risks of contamination when doping but it's a doable task. This was already accomplished by the research team I referenced earlier who created vertical tunneling transistors. The research proved that graphene could be used for logic albeit with a lot of jerry rigging. That doesn't mean graphene is a realistic silicon alternative. That's why other 2D materials that are actual semiconductors are researched in parallel and touted as having actual potential like platinum sulfide. I would use the analogy of nuclear fusion research which is pie in the sky research while thorium salt reactors are becoming a proven commodity soon to be commercialized.
I have a theory that IF China can commercialize quad patterned 7nm DUV process, that it will collapse margins for every process node below 7nm including the leading edge. Sometime after 2026, the excess semiconductor capacity that will be laying idle combined with what will probably become a largely self sufficient China semiconductor market above >=7nm, along with China's supply chain cost advantages, will totally destroy the global semiconductor business model. The price/performance curve between what will become China's DUVi 7nm chips and all leading edge EUV performance chips, will not warrant the price discrepancy that currently exists between the leading edge processes and processes 3 generations older. Imagine having chips performing at say ~50% of the leading edge, consuming say 200-250% the energy, but costing 5-20 times less.
