21
votes
TSMC may have approval to create 2nm chips in the US
Link information
This data is scraped automatically and may be incorrect.
- Title
- TSMC Might Have Secured Taiwan's Approval To Make Advanced 2nm Chips In US
- Authors
- Ramish Zafar
- Published
- Jan 13 2025
- Word count
- 536 words
TSMC has been angling for an approval to create their most fancy 2nm chips outside of Taiwan. The US is investing a lot into the local chip production and spends billions on acquiring their own foundries, notably the Arizona foundry, and is of course very interested in creating the most up to date chips locally.
The news isn't that TSMC is planning this, this was already a story in November '24 where the idea was to have 2nm production in the US start in 2028, but they did not have approval yet to create these chips in foreign countries. It looks like they may now have, or shortly receive, the approval to go ahead with that plan.
I worry that Taiwan erodes their leverage every time chip production is created out of country. That may not happen overnight, they still keep R&D in Taiwan, but having the most up to date production elsewhere may spur innovation elsewhere too.
Similarly to Russian imperialist expansion being destabilizing, Chinese expansion would be catastrophic for world stability too. Nuclear powers being allowed to take land is a dangerous game. One country doing so is a bad decision, two is a trend.
Tags: Politics may need to be added if the conversation steers in that direction. I don't intend to focus too much on China, but since chips are the oil of the 21st century it felt prudent to at least mention that angle.
I agree, but I suspect Taiwan's loss of leverage is strategic realism on the part of their leaders - they know that if China invades they'll have to flee, and not standing in the way of the US securing advanced chip supply chain means they're more likely to have a place to go.
I don't think China's invasion of Taiwan can be realistically stopped. For the US & allies Taiwan is just so far away for force projection on a scale needed to combat the scale of China's military build-up (especially given their much improved energy security: nuclear and solar build-up, and an increase in land-based gas pipelines to reduce vulnerability to a blockade of tankers moving through the Strait of Malacca).
Due to chip sanctions, China have been growing their own chip foundry industry as fast as they can, so there's also a strong economic incentive to invade while the world is dependent on Taiwan: if TSMC foundries are destroyed (or, worst case, under PRC control) that means realistically less competition for Chinese foundries even though they're currently worse than TSMC's state-of-the-art.
I wouldn't be quite so sure. It seems quite possible to me for Taiwan to make the strait very dangerous to cross. It's 160km wide. So 3-5 hours by ship. Plenty of response time for modern militaries. With the way Taiwan is structuring their military, I'd expect that any attempted crossing of the strait would run into a barrage of anti-air and anti-ship missiles. China would have to expect massive casualties. A "3 day special military operation" level of success wouldn't be surprising to me. It's certainly not a foregone conclusion in my mind. I hope Xi Jinping doesn't test my hypothesis.
It entirely depends on whether or not Xi allows that level of casualties and whether or not the US decides to uphold their defense agreement. If the US pulls out I do not doubt that China can take Taiwan.. eventually. China is far too large and not hurt by sanctions as much as Russia, but it will be incredibly costly. What we know is that Taiwan is very defensible but nothing is impenetrable. 160km is a dangerous trek for ships but it allows for air sorties from Chinese soil subsequently reducing their reliance on a carrier fleet and can just rely on a fleet in being and bombing raids. Air supremacy is a given with their J-20 fleet.
China is not leaving it to doubt: They openly state their goal is to take Taiwan.
Though should it happen I can't see any way that Taiwan will not fully sabotage their foundries.
All just speculation of course and something I don't wish to find out.
The US would have no choice but to intervene. We've known for a while that the centralization of chip production in Taiwan is a huge vulnerability for the West, and if it were threatened, it would be worth a whole ton of resources to keep it.
The consequences of cutting off chips to NATO countries would be absolutely disasterous, and I have no doubt that trillions of dollars and thousands of lives would be viewed as an acceptable price to pay to prevent that from happening. It's completely unprecidented in the history of the world that such a singular, small place in the world has the amount of global geopolitical, economic, and military strategic importance as Taiwan currently does.
As much as the US tries to sell it that way, the defense agreement absolutely isn't for Taiwan benefit. It's for the US's.
You'd say so, but that's exactly what I'm alluding to with this post. If the US gets the 2nm TSMC foundry in Arizona they are at tech parity with Taiwan -with a mere three year R&D lag-, vastly undermining the security risks that centralization now hold.
If the US no longer needs Taiwan to produce 2nm (or better) chips, I worry that under certain leadership the US will cease to uphold that defensive pact.*
The US is throwing billions upon billions to both TSMC and grants for Intel to become competitive with Taiwan, likely with the goal to not only decentralize production, but become the industry leader instead. You rightly point out that the defense agreement is to the US' benefit, but what if it no longer is?
*I do not think the US will abandon Taiwan if China declares while the defense pact holds, but as soon as it's no longer needed to protect chip production I can foresee the US slowly extricating itself from that pact before it gets hot opening Taiwan for that incursion. Similarly, as soon as it was clear the UK wasn't going to contest Hong Kong, China took the opportunity to move in well in advance of the lease expiring.
TSMC is only 1-3 years ahead of the US at this point. It'd be painful for a while as the capacity drops, but capability wise the world would catch back up soon enough. There's a chance Intel is competitive with 18A in 2025/2026 and possibly even pulls ahead of TSMC with 14A. Even if not, the world would adapt by just sidegrading chip capabilities for a few years. Not to mention that TSMC has fabs outside of Taiwan and competitive enough fabs.
China would also hurt from the destruction of TSMC and would hurt harder than the US would since their fabs are much more behind. Through that lens, does the US really need to defend Taiwan? It just needs to destroy TSMC if Taiwan were to imminently fall.
Taiwan needs to get more serious about defending against China with asymmetric weapons and stop wasting billions on naval vessels that China would instantly sink.
I somewhat agree with this statement solely because of how terrible of an employer TSMC is. There are many articles confirming this (here's one), and I believe they're trying to justify bringing in as many citizens of Taiwan as possible by claiming US citizens don't have the appropriate work ethic to be employed by them.