Hacker News Reader

Comment by gjsman-1000 | Hacker News Reader

To me, I think that Apple's reliance on China, and Taiwan, was and is an extremely risky and misguided decision. Imagine if Apple had developed a total reliance on chip factories built in, say, Ukraine, back in 2018. That's about where we are now with Taiwan and China, I believe.

It's sure been profitable - but at, I would argue, obscene risk. A few years of TSMC being unavailable would have Apple on life support. Apple would actually be worse off than many companies and the ditching Intel decision would do a total 180 into a massive mistake. Nobody cares if you have higher-performance products if you can't make them.

Edit: For those saying that can't happen, think Ukraine in 2012 or 2018. Russia? Invading? Can't happen. For Taiwan, China, invading? Can't happen.

View on HN

> For those saying that can't happen, think Ukraine in 2012 or 2018. Russia? Invading? Can't happen?

No one paying even a little bit of attention thought “can’t” since the invasion of Georgia in 2008. In 2012 they would have thought “has no reason to at the moment”, much as they would about Belarus today; what was unexpected in 2014 was the Maidan Revolution tossing out the Russia-aligned government, to which the invasion was an immediate response.

And in 2018 the likelihood of further invasion was widely discussed, while the scale and precise timing may have been unpredictable in 2018, that further invasion was likely, especially an attempt to seize the Donbas, where Russia has been supporting proxy forces since the 2014 invasion, was not.

> For Taiwan, China, invading? Can't happen.

That’s... Not at all the general thought.

> No one paying even a little bit of attention thought “can’t”

Maybe that's true in the abstract, but how did those in power act based on this obvious threat?

Europe completely refused to even begin to diversify its natural gas sources until this year, even though that obviously should have happened after the annexation of Crimea at the very latest. They invited catastrophe by relying on Russian energy.

In practice, they didn't take the threat seriously. They thought it wouldn't happen, they gambled quite a lot on that outcome.

> Maybe that's true in the abstract, but how did those in power act based on this obvious threat?

Again, it wasn’t an obvious threat, not because Russia lacked capacity or will but because it already controlled Ukraine via a puppet government. That changing, and very quickly from the emergence of a significant protest movement, was the unpredicted wildcard.

> Europe completely refused to even begin to diversify its natural gas sources until this year,

Baltic Pipe, a major component of the diversification effort, was identified as an EC interest in 2013, completed a feasibility study in 2016, had capacity agreements in place in 2018, and went online this September.

Infrastructure isn’t something that pops into place when you snap your fingers.

All of that being said, I wrongly predicted that Russia wouldn't invade, because I rightly predicted the invasion would be a disaster for Russia.

A CCP invasion of Taiwan would be even more disastrous for the PRC. They would need to conduct an amphibious assault, gain air superiority against "the quad" (no easy task), and would be unlikely to gain control of TSMC.[0]

I think my reason for miscalculating the Russian invasion had to do with reasons unrelated to the actual military campaign: Putin felt that Russia's culture was existentially threatened by the West, its ability to invade was monotonically decreasing, and believed that a largely ineffectual, predictable US administration that appears to follow the mantra "speak loudly and carry a paper straw" wouldn't resort to extreme measures to stop Russia. In other words, invasion made total sense from a certain political-cultural calculus.

I'm much more worried about the CCP, who have become increasingly ultra-nationalist, including concentration camps and racial-superiority indoctrination in schools (including something like a "stabbed in the back" narrative). The political crackdowns have become quite severe, and the Chinese financial is in bad shape, so an attempted invasion of Taiwan may increasingly be the thing the Party sees as bolstering their position domestically.

The invasion would still be a disaster, and I have a feeling that the US & Taiwan already have plans to relocate the personnel from Taiwan's TSMC facilities to the US in order to use their specialized knowledge in order to keep producing superior chips.

[0] https://www.jonstokes.com/p/why-a-chinese-invasion-of-taiwan

> > For Taiwan, China, invading? Can't happen.

> That’s... Not at all the general thought.

Indeed. AFAIK the general sentiment is that the only way they don't invade some time this century is if they manage to Hong Kongize them to transition them toward full annexation, through soft power.

That's a nice idea in theory. As much as I'm harsh on Apple for their decisions elsewhere, TSMC is the right choice, right now. Two things that are hard to ignore:

1. Apple held moving the Mac line on Intel for years despite clear indications that it was hurting them.

2. Intel has begun moving to TSMC to fab chips in order to be competitive with AMD who already moved.

What alternative move should Apple make? Move to Samsung, where a more expansionist China is still a problem? Start doing their own fabrication in the western hemisphere, a process that could take decades to complete.

It is likely that Apple wants to use proposed TSMC or Intel fabs in Arizona, but cannot today.

That's like super naive.

First, you mentioned it. "It has sure been profitable". That's kinda the whole point. You get more profits, you use those profits for R&D to further distance yourself from the competition.

If there is any way to reduce your risk it is building a technological edge over competition and sizeable cash reserves.

Second, even if Apple faced complete stoppage in production, its situation is different from other companies. It not only can do things that other companies can't (like pull cash to build entire supply chain in some other place), it also has devout user base that will mostly just wait for Apple to come back. People who buy say Samsung laptops and phones will switch to another brand with no second thought if they can't buy Samsung, while people who buy Apple will try as much to wait and will not switch easily.

Third, the US is currently the guarantor of Taiwan's safety so maybe the risk isn't as big as you think.

Fourth, if TSMC is unavailable they will spin the production somewhere else. And who do you think will get the first chips out of it -- the companies that are the most ready (Apple, they already have all the know how and designs), the companies that are ready to pay the most for the privilege (Apple, because of their cushy margins) and the companies that have the most weight behind them (again Apple).

So no, I think that saying Apple's risk is obscene is just naive. Apple has positioned itself to be best at weathering any storm that might happen, it knows the competition will be hit even more and is using this to reap as much benefits as it can while the party lasts.

> Fourth, if TSMC is unavailable they will spin the production somewhere else.

Where?

Intel would sell them their fabs, but they don't have a process as advanced as TSMC 5nm. Samsung might be able to compete, but their yields are wickedly low and still slightly behind TSMC. GlobalFoundries might accept Apple as a customer, but then we'd basically have two Qualcomms.

What is Apple's plan-b for Taiwan?

Isn't TSMC building outside Taiwan, to your point?

https://www.tomshardware.com/news/tsmc-fab-21-arizona

It is only 20000 wafers per month on 5nm. Only 5nm factories in Taiwan make 150000 wafers per month.

https://wccftech.com/tsmc-boosts-5nm-production-to-150000-wa...

If China invades Taiwan, it will not be possible to do business with China or Taiwan and every company with substantial exposure to this risk should be preparing for the possibility that all resources in China are instantly "lost" or unavailable for many years.

Apple's production capacity has taken decades to build and will take years at minimum to restore. Their exposure is so immense that it would make sense for them to spend a percentage of their cash position just on lobbying to try and prevent the outbreak of war.

Their risk is obscene. They're poorly positioned because their manufacturing base is undiversified and in the case of "complete shutoff of goods from China" it will be years before they're back at capacity.

A few years of TSMC being unavailable would cripple consumer electronics in general. Lots of chip vendors all buy their dies from TSMC. It's not just Apple by a long shot.

This is a known problem, potentially one even at the scale of being a national security issue. It's a big part of why there are all these new federal subsidies to try and bring advanced chip-fab back to the US.

There was no alternative at the time regarding overall production. China gave them the option to be extremely flexible and efficient. Other countries like India/Vietnam is copying the Chinese cookbook.

Ony TSMC is capable of producing the most performant chips.

> There was no alternative at the time regarding overall production. China gave them the option to be extremely flexible and efficient. Other countries like India/Vietnam is copying the Chinese cookbook.

Yes, and no. Apple has done manufacturing in America before, a long time ago (remember the Apple II?). Also with the $200B in cash they had, they could probably build every factory down to the screw factory they needed. Also worth remembering is that iPhone bodies don't change very much (iPhone SE is like, a decade of a design and still going). $200B would buy at least 5 or more latest-gen chip plants, easily, or at least give TSMC all the capital they needed.

If I was Tim Cook, I would have absolutely looked into segmentation. American manufacturing for American customers, Chinese manufacturing for European/Asian customers, Indian manufacturing for Indian customers.

> Also with the $200B in cash they had

When exactly are you talking about?

> $200B would buy at least 5 or more latest-gen chip plants, easily, or at least give TSMC all the capital they needed.

$200B and a decade. Also do you think Taiwan is just going to let go of their biggest bargaining chip for a measly $200B?

> If I was Tim Cook, I would have absolutely looked into segmentation. American manufacturing for American customers, Chinese manufacturing for European/Asian customers, Indian manufacturing for Indian customers.

I mean they are doing this as per the article.

> $200B and a decade. Also do you think Taiwan is just going to let go of their biggest bargaining chip for a measly $200B?

Not necessarily a decade. Intel is promising ~4 years for their fabs, and hoping to be online by 2025-ish if I recall correctly. Fab production is also, it seems, going quicker than expected.

> When exactly are you talking about?

In 2018, Apple's cash pile hit $285 Billion. I was underestimating it. Imagine if they had started building fabs back in 2018 or earlier - they would be coming online, potentially, right now. Which would have been great.

> I mean they are doing this as per the article.

I'm looking at geopolitics and thinking this could end up being too little too late.

https://www.cnbc.com/2018/02/01/apple-earnings-q1-2018-how-m...

Your point is very strange. Apple did not just suddenly decide in 2018 to manufacture in China. It is what they have been doing for like 10-15 years before that. If you count from iPod on. In mid 2000s it certainly was not an option to manufacture just anywhere and certainly not by them.

> Not necessarily a decade. Intel is promising ~4 years for their fabs, and hoping to be online by 2025-ish if I recall correctly. Fab production is also, it seems, going quicker than expected.

Ok and how many generations is Intel behind TSMC now and how many will they be in 4 years?

You think Apple investors and customers are suddenly going to be OK with a new version of the iphone being worse than its predecessor? To be feasible they can't just catch up, they have to surpass.

Intel already has plenty of fabs in the US. There are also a number of semiconductor companies with fabs in the US. They might not be as advanced as TSMC, but the infrastructure and workforce is already in place. What is lacking is a desire to dump a bunch of capital into fabs in an industry that has a history of severe boom/bust cycles.

> In 2018, Apple's cash pile hit $285 Billion. I was underestimating it.

Much has changed since 2018.

Apple cash on hand for the quarter ending June 30, 2022 was $48.231B, a 21.82% decline year-over-year.

Apple cash on hand for 2021 was $62.639B, a 31.12% decline from 2020.

Apple cash on hand for 2020 was $90.943B, a 9.56% decline from 2019

Apple cash on hand for 2019 was $100.557B, a 51.67% increase from 2018.[1]

[1] https://www.macrotrends.net/stocks/charts/AAPL/apple/cash-on...

Yeah but they probably spent the money on stock buybacks.

Believe it or not... cat toys. Accounting is still investigating.

Apple went from the brink of bankruptcy to a $3T market cap, all on the back of Chinese/Taiwanese manufacturing. It's pretty safe to say that the move wasn't "misguided".

There are a lot of companies dependent on TSMC, not just Apple.

I don't dispute that - but Apple is still on the line the most of any company if TSMC were to become unavailable (in sales, in products requiring TSMC chips, lost sales, Apple Silicon on all of their devices, so forth). Also Apple had the most capital of any company and thus the least excuse for such dependency. 4 years ago, according to CNBC, they had $285 billion in cash reserves. That could've definitely bought some fabs in America that would be coming online right now.

NVIDIA still can make parts in Korea if need be, even if they are a generation behind. AMD could be manufacturing on Intel foundries before we know it. Samsung has Exynos in Korea as well, which is better than nothing.

I don't see China invading Taiwan, since the U.S. has made it abundantly clear that they'll intervene militarily, and the Chinese know they are (as of yet) no match for the U.S. They'll bide their time. Chinese think in terms of centuries, not decades. The hive's important, not the individual worker.

I do think they'll need to wait a very long time since if the West decouples they'll be thrown back decades if not more, with no hope of ever catching up.

But if a spat were to occur, Apple would definitely be hurt...BIG TIME. In fact, I could even see the company collapse if it were to be cut-off from its supplier base in China and Taiwan.

So the alternative is Samsung fab located in South Korea. Technically Korean war isn't ended and North Korea actively developing missiles.

> For those saying that can't happen, think Ukraine in 2012 or 2018. Russia? Invading? Can't happen.

They "invaded" Crimea in 2013, a lot of people expected what's currently happening.

edit: 2014 not 2013. Also Crimea overwhelmingly supported being annexed into Russia as they were severely mistreated by Ukraine.

There were very clear tensions leading up to both Crimea and Ukraine events.

But using the political events of Ukraine/Russia to support tossing Taiwan to the PRC is not something I agree with.

Incorrect. Crimea was invaded in 2014. My point was that in 2012, it was 2 years before the invasion, and less foreseeable despite how quickly in the grand scheme of things it came afterwards.

Additionally Ukraine is the breadbasket of the world, people thought Russia wouldn’t do this because of how it would mess up the food supply chain… the exact same argument being made for why China can’t invade Taiwan, too many companies rely on it.

> Additionally Ukraine is the breadbasket of the world, people thought Russia wouldn’t do this because of how it would mess up the food supply chain…

No, they thought Russia wouldn’t do it because Ukraine had a Russia aligned government up until just before the invasion.

The Russian armed forces has had a lot of actual battlefield experience invading Georgia in the ‘00s, bombing Syria in the ‘10s, not to mention its internal bloody counterinsurgency occupation in Chechnya in the ‘90s. The Chinese military hasn’t fought a real war since against Vietnam in 1979, and its navy has had no experience at all since- Korea? The end of the civil war?

War is entirely more uncertain for China, and potentially more destabilizing with more to lose, than it was for Russia.

The PRC constantly skirmishes with India and has land grabbed many islands in the South China Sea, many already occupied.

Yes, since and during the communist takeover of China, all of its military might has generally been directed inwards at it's own people, but they are still a military force and are well organized.

I'd equate the current period of the PRC to the Soviet arms build up of the Cold War.

Yes they have a large and considerably upgraded military, but border skirmishes and planting flags on unpopulated rocks are hardly comparable to Russia’s recent post-Cold War military experience. The point is how the PLA would fare in an actual war is far from known yet.

As far as the Soviet military went, the fallout of that buildup, and the actual condition of the Red Army, are known after the fact now.

> Yes they have a large and considerably upgraded military, but border skirmishes and planting flags on unpopulated rocks are hardly comparable to Russia’s recent post-Cold War military experience.

Skirmishing against a near-peer like India is probably better preparation for a large scale war against a well-financed enemy with modern equipment, doctrine, and Western support than Russia’s wars against lesser former Soviet states and in support of Syria.

Were they actually "skirmishing" though? I thought it was mostly border guards scuffling, even engaging in hand-to-hand combat at one point.

What we’ve found out in Ukraine is that much of the military spending has been siphoned off by corrupt leadership. How do we know that the same thing isn’t happening in China?

> What we’ve found out in Ukraine is that much of the military spending has been siphoned off by corrupt leadership

We didn’t find that out in Ukraine, we’ve known that from fairly widely circulated news, often from direct participants in the corruption (such as the people to whom essential gear was sold), consistently since at least the First Chechen War of the mid-1990s (where sales of fuel, weapons, and ammunition by Russian military personnel with custody over them to the enemy were a significant and recurring thing.)

What we have learned in Ukraine is a better idea of the extent of the effect of that longstanding problem on current Russian military capacity.

Now, it could be that China has a similar problem and has been better at keeping it under wraps than Russia has been for the past 30 or so years, but there is not much reason to suspect that.

RuAF war experience mostly involved insurgents, or very weakened states (Georgia 08, Ukraine 2014). Ukraine 2022 seems to be a much tougher nut to crack. Even though by now the Russians have 8 more months of intense battlefield experience, they seem to be worse off strategically than at the beginning of the war.

It is in Taiwan's interest to be at least as tough a nut as Ukraine is, and I think they are trying.

2014.

Reply on news.ycombinator.com