3 ms·

There’s also a difference subsidizing large companies that relies on being at the top of the supplier chain vs those that exist somewhere near the bottom or the

by 1-6 5y ago

There’s also a difference subsidizing large companies that relies on being at the top of the supplier chain vs those that exist somewhere near the bottom or the middle. Evaluating the tech stack is very important and usually at the very top you find trivial technologies that are not difficult to replicate. This is where you find brand, trademarks, and marketing.

moralestapia 5y ago
>usually at the very top you find trivial technologies that are not difficult to replicate Maybe in other markets, anything related with semiconductors has a HUGE capital barrier to entry. Exhibit A: ASML's EUV machines.
- wincy 5y ago
  I think the concept GP was using is lowest level of abstraction to highest level. Twitter can be cloned easily. ASML’s EUV machines cannot. I don’t think he meant “high technology”
  - pmontra 5y ago
    By the way, a common pattern of technology trasfer is a company buys a product, reverse engineer it and in a few iterations (which could be many years long) it competes on price or on quality with the original manufacturer. I understand that an EUV machine is much more complex than a chisel but how difficult (in time) would be to clone one? Did anybody already start doing it? A scenario is: China invades Taiwan, the chip factories blow up so China doesn't get them, it takes years to build them somewhere else for the rest of the world but China won't have those machines from ASML ever (embargoes.) But can they buy one now and clone it in a reasonable time?
    - tooltalk 5y ago
      My wild speculation: without any outsider's help, 30-40 years, or most likely never. China has barely 10 years of experience in chip industry and it's not necessarily the most profound one. China's most successful foundry SMIC was made possible by former top TSMC engineers (many with PhD's from top US universities) and it's generally understood that SMIC's most cutting edge 14nm node is really a clone of TSMC N16FFC. Many also suspect 14nm is their limit and can't advanced any further without access to the chip manufacturing supply-chain in the US/Japan/EU, who collectively own 95% of the market. Shanghai Micro Electronic Equipment, SMEE, recently announced that they developed their own 28nm lithography machines, but that's also with help and components from Japan. The 28nm processing is hardly a cutting edge tech and that's not even under US sanction. Even Taiwan/South Korea depend almost entirely on supply-chain from the US/Japan/EU (except wafer). Now, the key idea here is "outsider's help." It's not impossible if China is allowed to poach engineers from their suppliers/competitors, steal IP, or import highly sensitive, cutting-edge components and materials from the US/JPN/EU. The problem as I see it is, there is obviously no shortage of American multinational companies, such as Apple, Nike, etc,... willing to sell out America's national security or economic interests to make more doh's in China. And Apple isn't just a very profitable US company with the largest market cap in the world, but also it's very well politically connected -- just look at the shenanigan Apple pulled against Qualcomm with their operatives at FTC. The orange man was somewhat of an anomaly and managed to stop this flow, but I'm not so sure about Biden or whoever comes next. The recent revelation, or $270B deal between Tim Apple and Xi, clearly shows what Apple intends to do and where this is going to go if Apple is left unchecked.

by 1-6 5y ago

moralestapia 5y ago

>usually at the very top you find trivial technologies that are not difficult to replicate Maybe in other markets, anything related with semiconductors has a HUGE capital barrier to entry. Exhibit A: ASML's EUV machines.

wincy 5y ago
I think the concept GP was using is lowest level of abstraction to highest level. Twitter can be cloned easily. ASML’s EUV machines cannot. I don’t think he meant “high technology”
- pmontra 5y ago
  By the way, a common pattern of technology trasfer is a company buys a product, reverse engineer it and in a few iterations (which could be many years long) it competes on price or on quality with the original manufacturer. I understand that an EUV machine is much more complex than a chisel but how difficult (in time) would be to clone one? Did anybody already start doing it? A scenario is: China invades Taiwan, the chip factories blow up so China doesn't get them, it takes years to build them somewhere else for the rest of the world but China won't have those machines from ASML ever (embargoes.) But can they buy one now and clone it in a reasonable time?
  - tooltalk 5y ago
    My wild speculation: without any outsider's help, 30-40 years, or most likely never. China has barely 10 years of experience in chip industry and it's not necessarily the most profound one. China's most successful foundry SMIC was made possible by former top TSMC engineers (many with PhD's from top US universities) and it's generally understood that SMIC's most cutting edge 14nm node is really a clone of TSMC N16FFC. Many also suspect 14nm is their limit and can't advanced any further without access to the chip manufacturing supply-chain in the US/Japan/EU, who collectively own 95% of the market. Shanghai Micro Electronic Equipment, SMEE, recently announced that they developed their own 28nm lithography machines, but that's also with help and components from Japan. The 28nm processing is hardly a cutting edge tech and that's not even under US sanction. Even Taiwan/South Korea depend almost entirely on supply-chain from the US/Japan/EU (except wafer). Now, the key idea here is "outsider's help." It's not impossible if China is allowed to poach engineers from their suppliers/competitors, steal IP, or import highly sensitive, cutting-edge components and materials from the US/JPN/EU. The problem as I see it is, there is obviously no shortage of American multinational companies, such as Apple, Nike, etc,... willing to sell out America's national security or economic interests to make more doh's in China. And Apple isn't just a very profitable US company with the largest market cap in the world, but also it's very well politically connected -- just look at the shenanigan Apple pulled against Qualcomm with their operatives at FTC. The orange man was somewhat of an anomaly and managed to stop this flow, but I'm not so sure about Biden or whoever comes next. The recent revelation, or $270B deal between Tim Apple and Xi, clearly shows what Apple intends to do and where this is going to go if Apple is left unchecked.