Hacker News Reader

Comment by cronix | Hacker News Reader

For the sake of argument, let's just say that tomorrow Taiwan is suddenly taken over by another country and TSMC is no longer an option. Everything flowing from Taiwan just stops, despite agreements or money spent. That becomes null and void. I'm not sure what good cash on hand will do them within x years of that event happening, should it. It seems Apple is putting more and more of their eggs in TSMC's basket and it seems TSMC is the only company who can get Apple where it wants/needs to be. At least currently. From what I can tell Apple is totally coupled and reliant on TSMC at this point. Only a few of the upper end product lines are still using Intel processors and you know they'll soon be gone, but they are a very small part of Apples total market share by revenue at this point anyway.

View on HN

> despite agreements or money spent. That becomes null and void.

the only way to prevent this from affecting you as a corporation is to have enough military power to prevent the scenario from happening...oh wait, that's what the US military is for.

That is true. It's really hard to find news in MSM about this stuff though. I had to get to page 3 on google (tried ddg first!) results to bypass defense news type sites as most people would just poo-poo those. Things haven't been looking so good. https://www.nbcnews.com/politics/national-security/china-s-g...

Japan takes a stance: https://www.npr.org/2021/07/26/1020866539/japans-position-on...

Issues in Philippines: https://www.bbc.com/news/world-asia-56474847

Things have been heating up in the region, for the last few years especially, and generally not trending in a direction I think most of the world wants to see. I don't think the scenario I proposed is that far fetched, and the probability is continually trending is a bad direction. China is a rising super power and by most accounts will rival/surpass the US in many ways, very soon. Within a decade.

One way to accelerate that is to throw a monkey wrench in the world economy while you are largely insulated from the results. At least more insulated than your competitors are. I wouldn't even say that hitting a few Samsung fabs in the process was out of the question. It's not like the response will be much worse than if they didn't touch Samsung and only took Taiwan. Might as well?

I hate thinking this way and I could very well be wrong in any or all of my arm-chair analysis. I absolutely hope so. China's stance on reuniting Taiwan back with the mainland has been growing as Taiwan's power has been growing.

> BEIJING/TAIPEI (Reuters) -Chinese President Xi Jinping pledged on Thursday to complete "reunification" with self-ruled Taiwan and vowed to "smash" any attempts at formal independence, drawing a stern rebuke from Taipei, which lambasted the Communist Party as a dictatorship.

> China, which considers democratically-ruled Taiwan its own territory, has stepped up efforts under Xi to assert its sovereignty claims, including regular flights by fighter jets and bombers close to the island.

> "Solving the Taiwan question and realising the complete reunification of the motherland are the unswerving historical tasks of the Chinese Communist Party and the common aspiration of all Chinese people," Xi said in a speech on the 100th birthday of the ruling Communist Party.

> "All sons and daughters of China, including compatriots on both sides of the Taiwan Strait, must work together and move forward in solidarity, resolutely smashing any 'Taiwan independence' plots."

https://news.yahoo.com/chinas-xi-pledges-reunification-taiwa...

> I'm not sure what good cash on hand will do them within x years of that event happening

Extradite key people from Taiwan? Buy up ASML's inventory and next N months' production at an exorbitant premium?

Or, just weather the storm. Apple would be hit. But so would everyone else. I'd wager that hit could take out some of their competition.

The competition largely either has its own supply like Samsung or is owned by/from that other country (Xiaomi, Huawei, Oppo, Vivo, Realme, Motorola).

I guess Samsung would benefit greatly.

It’s hard to imagine a situation that both removes Taiwan from play and preserves peace in Korea. It’s unfortunate the world’s chip supply is so geopolitically concentrated.

PRC invades ROC/Taiwan. Besides some saber rattling, US, Japan, Korea don't get involved. The end.

Is that unthinkable?

It's not unthinkable, but continuing this line of thought, is it unthinkable that China makes a point of upholding all contractual obligations leaving TSMC's customers unaffected?

Just dropping some interesting posts I found when researching this general area earlier today:

https://unherd.com/2021/07/will-china-invade-taiwan/

Will China invade Taiwan? A panel of superforecasters considers the likelihood of an imminent global conflict

https://www.jonstokes.com/p/why-a-chinese-invasion-of-taiwan (Jon Stokes, the Ars Technica founder)

Why a Chinese invasion of Taiwan would be a catastrophe for China and the world. If they do it, it won't be for the microchips

I believe that PRC doctrine for an invasion of Taiwan is to launch preemptive strikes against US bases in the region. So unless they change their strategy before they invade, the US will definitely get involved.

Who would have an interest in drawing South Korea into it?

The US would be primarily responsible for forcing Korea into the war. The border of North and South Korea is only about a day drive from Beijing, allowing the US to threaten the Chinese government from the very beginning of any conflict. The situation on the Korean Peninsula is commonly thought of as a failing for US diplomacy, but the conflict actually provides perfect cover to keep a huge military detachment almost on the border of our chief geopolitical rival. It’s highly advantageous to the US for Korea to remain fractured.

> Buy up ASML's inventory and next N months' production at an exorbitant premium?

Which inventory? As far as I know ASML is already churning out as many EUV machines as they can and yet they are still fully booked for the next few years. They're currently limited by personnel, factory space and their own supply chain.

Apple would take a hit in that case, but so would everybody else who use TSMC. Intel wouldn't be hit, but they also don't make ARM chips.

Apple makes a premium product, which means that they can outcompete others, especially Android, for the remaining chips.

Even in the event that there is no new iPhone for the year they will still make money on their services, and a lack of iPhone will not make it less desirable. Other smartphone makers won't be able to stay in business.

Android phone makers won't be as affected since midrange/lowend devices can use older manufacturing processes which are (relatively speaking) widely available.

Apple needs the best to keep its edge and justify the premium.

Well, most of (a while back it was "almost all of"... it might still be) the profit in Android is made by Samsung, which makes high-end devices that compete with the iPhone.

That is a good point, but I wonder: how many second-rate chips are available and how bad are they? Because if the top end is 3 nm and they need 10 nm, then I don't know if there are enough other places to get them.

I think Android phones aren't really CPU/GPU constrained anymore. I bought a cheap Realme phone (130 EUR) for a relative and the thing runs everything you throw at it fluidly, even pretty demanding games. (SOC is MediaTek G95 on 12nm).

TSMC is building a new fab in Arizona as a hedge for this scenario.

That fab will do 5nm in 2024, I think Taiwan TSMC will be on < 3nm by then! https://www.datacenterdynamics.com/en/news/tsmc-starts-work-...

And is thinking about building another one in Europe.

Antifragility is not absolute. Stress your muscles, you'll grow strong. Overload them by trying to catch a falling elephant, you'll die. Taleb says as much in his book.

The unique property is that the system can get objectively better from being stressed, not that it is a superhydra.

Well I use my iphones for about 5 years so they could continue funding salaries from cash while getting new supply online.

> For the sake of argument, let's just say that tomorrow Taiwan is suddenly taken over by another country

How about let's not? There are infinite possibilities you can imagine of things going wrong. Pulling one example out of an infinite list does nothing to address the argument itself, it's just fictional role play. And every time one plays a scenario out, they could always just go one more step towards a more catastrophic scenario, until we'd arrive at "A-ha, but as we see, Apple's strategy fails to address heat death of the universe! Got 'em!" Of course your suggested scenario isn't as absurd, and the global supply chains reliance on a few privately held companies is indeed problematic. But the discussion was about Apples strategy and how they prepare with more foresight than their competitors, succeeding better under pressure. Not that they are omnipotent gods that can control and prepare for everything.

> Pulling one example out of an infinite list does nothing to address the argument itself, it's just fictional role play.

That’s true, but as a Taiwanese I can tell you Taiwan being taken over by another country is definitely in the shortlist of plausible scenarios. Not particularly likely, but definitely not fictional at all, maybe somewhere around Pearl Harbor possibility.

I suggest reading the whole comment. I address this concern directly in it.

Yes, there is a global supply chain fragility when it comes to a few companies controlling the supply of microchips. Nobody is suggesting that Apple somehow can do everything from scratch. Even if they had factories making the chips themselves, what if their raw material suppliers collapsed? What if a critical delivery route gets compromised, as happened with Ever Given? What if a war breaks out? What if, what if, what if... You can entertain any number of ideas. Everyone knows that the global supply chains have dependencies, and if something bad happens in the world, the economy and supply chains of the world suffer. This has nothing to do with Apple specifically.

I know you did, but it does not make the accusation less, uh, irritating to a Taiwanese, lacking a better expression (I don’t feel that strong, but do feel uncomfortable). The point is not how the threat can be addressed in what way, but you are greatly underestimating the possibility. Which is quite reasonable tbh, I don’t feel anything about wars in Africa and the Middle East and would likely underestimate how unstable things are when depending on entities in those areas, but it does not make the actual threat less worth considering.

Oh I see. I did not mean this as a political comment or an evaluation of the likelihood of the scenario playing out. Just pointing out that a single company cannot do much, if/when such things happen.

> There are infinite possibilities you can imagine of things going wrong. Pulling one example out of an infinite list does nothing to address the argument itself

You are wrong. This kind of example addresses the argument at its core.

Being "antifragile" is all about how you respond to low probability tail events such as this one.

If you become weaker as a result of such an event, you are not antifragile. You are fragile.

If you can withstand said event, you are not antifragile, you are just robust.

If you become stronger as a result of such an event, only then can you claim to be antifragile.

Haven't read "the antifragility book", but I'm still pretty sure antifragile does not mean immortal. The example in the article - muscles - grow when they are stressed, but they definitely can be destroyed also when enough stress is applied.

Or, if I am wrong, and indeed antifragile means immortal then I don't know why we are talking about antifragility rather than whether Apple is a company that cannot be affected negatively no matter what happens in the world. I'd also like to know why is there a need for a new synonym for immortality.

The book was written by Nassim Taleb and things like a global pandemic or supply chain collapse are exactly the kinds of things he talks about when he formulates idea such as antifragility.

Now, you are right: there's a limit to antifragility. I don't think the book clarifies this point.

But anyway, Taiwan being consumed by China is exactly the kind of thing that Black Swan and Antifragile deal with.

I mean, that's the whole point of the book.

Yes, but then you are talking about the fragility of the global supply chain system and capitalism. I agree, it is fragile, as is apparent with boom-bust cycles and things like covid or one ship getting stuck in a canal causing major disruptions. Not to mention how badly poor countries get shafted in the process. The global capitalist system is fragile, as history and current events show us.

But these problems are way, way larger than one single company. Are we really expecting Apple as a company within the global capitalist system to just isolate itself from the network and do everything themselves without relying on anyone anywhere? At that point no company in the world is antifragile, and never can be if it works within the capitalist system or relies on any sort of co-operation between people.

> Are we really expecting Apple as a company within the global capitalist system to just isolate itself from the network and do everything themselves without relying on anyone anywhere?

That's not really what it's about.

I would say an antifragile system is one that adapt to changes (even unexpected one).

A good example are businesses that were able to adapt to covid and earn more revenue by doing things like providing delivery services among other things.

I don't think antifragility is achievable by some kind of process. You probably need live players[0] at the helm to take the lead and figure out ways to adapt to new circumstances.

[0]: https://samoburja.com/live-versus-dead-players/

Reply on news.ycombinator.com