2 ms·

I assume that is if you already have the know-how. Starting from (almost) scratch would require more upfront work, right?

jvanderbot 5y ago
The US isn't starting from scratch. And presumably the total cost includes all the R&D cycles. TSMC was started by a Texas Instruments employee. But, if we assume they were, just double the cost as a sign-on bonus for all the engineers and poach away. This is problem that can be solved with money, and not a lot of it in the grand scheme of things.
- jasonwatkinspdx 5y ago
  You vastly underestimate the institutional knowledge necessary to operate state of the art fabs. Intel has been at the forefront of the field for decades, and even they are stumbling vs the current challenges. TSMC was founded over 3 decades ago. That it was founded by a TI employee says nothing about the situation today. The US is subsidizing a TSMC plant here, which should help the situation, as it'll give us access to their engineers and cross pollination. But it's still tiny compared to the need if we want to move the majority of our chip imports to domestic.
  - jvanderbot 5y ago
    I agree that I'm underestimating the startup cost. That's inevitable, I think. I can only counter by saying that I think we're also underestimating the benefits of being self-sustaining at the scale of TSMC, especially with the changing political landscape re: China.
dan_quixote 5y ago
This is a huge hurdle. There is actually some chip/silicon fab sites that never left (only scaled down considerably from their peak). A family member has worked at one for many years as an engineer/manager. When he tells me stories about the insane complexity of their process engineering it really hits home. Impurities at the nanometer scale can ruin 10 million dollar batches of silicon. Nasty chemicals are piped everywhere at the plant - the kind that explode if touching oxygen in the atmosphere or acids that dissolve bone before they noticeably affect your skin. Running such a production is difficult.

jvanderbot 5y ago

The US isn't starting from scratch. And presumably the total cost includes all the R&D cycles. TSMC was started by a Texas Instruments employee. But, if we assume they were, just double the cost as a sign-on bonus for all the engineers and poach away. This is problem that can be solved with money, and not a lot of it in the grand scheme of things.

jasonwatkinspdx 5y ago
You vastly underestimate the institutional knowledge necessary to operate state of the art fabs. Intel has been at the forefront of the field for decades, and even they are stumbling vs the current challenges. TSMC was founded over 3 decades ago. That it was founded by a TI employee says nothing about the situation today. The US is subsidizing a TSMC plant here, which should help the situation, as it'll give us access to their engineers and cross pollination. But it's still tiny compared to the need if we want to move the majority of our chip imports to domestic.
- jvanderbot 5y ago
  I agree that I'm underestimating the startup cost. That's inevitable, I think. I can only counter by saying that I think we're also underestimating the benefits of being self-sustaining at the scale of TSMC, especially with the changing political landscape re: China.

dan_quixote 5y ago

This is a huge hurdle. There is actually some chip/silicon fab sites that never left (only scaled down considerably from their peak). A family member has worked at one for many years as an engineer/manager. When he tells me stories about the insane complexity of their process engineering it really hits home. Impurities at the nanometer scale can ruin 10 million dollar batches of silicon. Nasty chemicals are piped everywhere at the plant - the kind that explode if touching oxygen in the atmosphere or acids that dissolve bone before they noticeably affect your skin. Running such a production is difficult.