Intel is dying a slow death, ARM is leaving UK's control, and TSMC is slowly taken over by China. The tables are turning.
The US companies do what they do best: innovate business models and find ways to vendor lock in and own the market.
But who's been doing all the research keeping moore's law alive? Lithography?
For a country as small as the Netherlands, you can wonder if we could afford from a geopolitical point of view to be producing all the world's chips (answer: no, it would force other place to invest highly in competition out of strategic interest).
Hence the strategy of being long term profitable and to just sell the machines to everyone at the same price.
It does make it baffling why Intel is unable to produce 5nn chips with the same ASML machines TSC is using. It must be management failing Boeing style.
As for the rest of Europe. Im sure they can politically do this. But where will the money end up?
It will end up in the pockets of friends and families of politicians that will be in "the management".
So don't expect anything from this. The future most sold chip will still be Apple Silicon or Samsung Exynos in TSC's foundry or Samsung foundry on a machine by ASML using IP from ARM and NVidia
AFAIK, ASMLs fancy EUV tech comes from a Bay Area company they acquired a while ago and not from the Netherlands.
I see reference for a company bought by ASML in California, HMI. They were bought in 2016 but the development of EUV technology predates it by many years so I find it unlikely that it had an important role in it.
Applied Materials - 17bn
ASML - 11bn
Tokyo Electron - 10.8 bn
Lam - 9.5 bn
KLA Corp - 4 bn
Cymer (2012, before acqusition) - 600m
ASML machines do not produce chips, they're just ones doing the etching. There is a ton around what you do with it, and how all the rest of the chip is produced.
They don't come with pre-programmed transistor gate templates. Each fab company has its own libraries of extremely specialized designs, doing different optimizations and tradeoffs. As mentioned plenty of times before, the consensus seems to be that Intel tried to pile up too many new improvements, which backfired. TSMC took a more incremental and safe approach. You have to appreciate that a defect that impacts high-volume yield today could have its roots in a design that was put on the roadmap years before.
> As for the rest of Europe. Im sure they can politically do this. But where will the money end up?
If H2020 is anything to go by, the majority of that money will be salaries of people working on big research projects. At worst the project outcome is useless. I don't see how any of the way these projects are structured benefit the pocket lining of managers.
Can you point to some industry literature that explains this?
If this is true, it is a lot more surprising that Intel is talking about going fabless. You build a lot of expertise by rolling the dice on new techniques that don't pan out on the first iteration. People wouldn't be so pessimistic about Boeing if they had just been overly aggressive in a new airframe design, rather than failing to safely mount new engines on an old design.
Well... yes. That's called judgement, and execs at firms like Boeing get paid positively stupid sums of money to provide the good kind. Which they didn't.
"What's the big deal? If I had left the vat of acid unattended in a locked room, everything would have been fine."
They’re not on the first iteration at this point. They’ve been on 14 nm since 2014. 10 nm and to a lesser extent 7 nm is failing after many iterations and many years.
Intel buying some TSMC capacity does not mean it is going fabless. Nor does it mean Intel is abandoning being vertically integrated.
I like your Boeing remark. It's better to fail at doing something aggressive than to fail to innovate.
I don't think that's a particularly useful way to divide up failures. I think it's better to fail in a way that doesn't kill people than a way that does. Whether you kill people by being too conservative or too aggressive with your design is pretty irrelevant compared to that.
The original article is here:
https://www.semiaccurate.com/2018/08/02/intel-guts-10nm-to-g...
It's paywalled but Charlie later published something else referring to that article as talking about COAG problems so I don't feel bad sharing the info at this point.
https://semiaccurate.com/2020/08/13/intel-talks-about-10nm-s...
Unfortunately Boeing doesn't have a huge list of customers that are willing to spend what it costs to switch to a new airframe. They sell to airlines what airlines want to buy. Airlines wanted a 737 with higher bypass engines and that's exactly what they got.
Boeing deserves what they get for pushing a faulty design out the door but they didn't do it as some nefarious cost cutting measure. They did it because American and Southwest said they wanted nothing to do with a "797" and threatened to switch to Airbus.
Maybe they are responsible for pressuring Boeing too much, but Boeing put defective airplanes on the market
so a nefarious moneygrabbing measure?
At the same time you can trust that TSMC secrets are no longer secret[2]. A lot of the Taiwanese companies are actually quite bad at security despite Taiwan have some top notch security engineers. I guess it's just old industries general problems.
[1]: https://bits-chips.nl/artikel/asml-is-too-convenient-a-targe...
[2]: https://www.extremetech.com/computing/313599-chinese-hackers...
100% agree. Considering the incredible education system and engineers we have, the 2 only things restraining Europe from reaching Silicon Valley level are corrupt politicians and risk-adverse VCs.
- labor laws (very hard to fire people in many European countries)
- lots of disjointed small-to-medium markets (with different laws and languages) make expansion on continental scale much harder than in the US
That's just not true now and hasn't been the case anymore since like the 2009 financial crisis. In most of Europe it's trivial to fire an underperforming non-unionized employee as the countries have adapted more business friendly laws to remain competitive after the financial crisis. Sure, there may still be some grey beards somewhere with old bulletproof contracts but for anyone entering the tech workforce now, getting fired is a rubber stamp basically especially since tech workers are not unionized and sometimes benefit from less protection than even some blue-collar professions which are unionized.
For example, in Austria you can fire anyone for no reason at all as long as you give them their notice period(usually 1 to 3 months) without them being entitled to any other financial compensation from their employer. Moreover, non-competes are legal here without any form of compensation and you can even fire employees while they're on sick leave. All this legally!
The whole you can't get fired in Europe maybe only applies to Scandinavia or France but it's mostly a meme nowadays.
Actually, contrary to popular belief, tech workers in California might enjoy better protection than tech workers in some European countries. Let that sink in for a bit.
On the other hand Silicon Valley is in California and has some of the strongest labor laws of the US, so if this is such a huge factor why is everyone in SV?
Even under the assumption that it's an important factor, would it be worth it for EU to adopt such policies considering the very real human cost?...
In the larger ones (DE, FR) labor laws are a nightmare.
France was famous for their approach of making an employees' life miserable in order to make them leave as they couldn't fire them easily. Honestly, I'd rather just be fired and get on unemployment than be subjected to psychological torture every day in a poisonous work environment.
Well they usually do it by compensating the employee quite heavily. Psychologically torturing an employee isn't that good for team morale, even if the employee was underperforming. I also don't think it's as culturally prevalent as you imply.
Where outside of the US/UK does that happen? Never heard in Europe that you get paid more so you leave when they can boot you out for free and I talked to lawyers about this. The reasoning behind the generous safety nets in Europe is also to make it easier for employee-employer terminations in case it doesn't work out. The laws here are that employers aren't allowed to abuse you, force you into unpaid overtime, etc. but fire you for free they can otherwise nobody would dare become an entrepreneur/investor in an environment where it's impossible to fire someone.
> when they can boot you out for free
You can't really. It's extremely difficult and no lawyer can magically change the fact that the culture and the courts are in favor of the worker, not the employer.
Now the laws might change a bit, making the terms of termination a bit easier for the employer in the future (we'll see what the next government brings), but in general I expect it to still be quite difficult. So it may be easier to pay the employee 30K extra compensation to let him go then trying a game of psychological torture. And if that sounds unreal to you, know that in the Netherlands employees can get paid up to 2 years if they have a nervous breakdown. And nervous breakdown is quite common culturally. So employees have dirty tricks of their own if the employer wants to play dirty.
Now of course, there are obvious disadvantages to this approach, not necessarily for the employer actually. The employer can protect himself by simply not hiring anyone for more than a couple of years, making "flex work" a real problem in NL. But yes, firing could be hard and I'd be surprised if it's much different in Germany/France etc.
As much as I like being on the side of the little guy, in any civilized society, the courts should be impartial and not favor the employer nor the employee and pass verdicts based on the law and not based on the culture and if the law says you can fire someone who continuously underperforms over a period of time then that law will be applied.
>After 3 years [...] a worker has to be moved to a permanent contract.
That sounds pretty poor IMHO. In most of the EU you're permanent after passing the first year of employment. I've never worked more than 3 years in a place so far, so in the NL I would have never reached permanent status, continuously being at the risk of on the spot termination. I dunno man, sounds pretty bad for me IMHO.
>And nervous breakdown is quite common culturally.
That actually sounds legit bad IMHO if mental issues are the becoming the norm in employment relationships.
>The employer can protect himself by simply not hiring anyone for more than a couple of years, making "flex work" a real problem in NL
I used to work for a top NL based company and this situation was often discussed around the office that NL companies are abusing expat workers from poorer countries by promising them riches in exchange for hard work and sacrifice and instead using them for 2-3 years then rotating them with new workers until they get fed up with this and move back to their home countries. Rinse and repeat. Sounds really good for Dutch business though.
Gonna ignore that one.
> I used to work for a top NL based company and this situation was often discussed around the office that NL companies are abusing expat workers promising them riches and instead using them for 2-3 years then rotating them with new workers until they get fed up with this and move back to their home countries
I really haven't experienced this at all. Tons of expats I know are on permanent contract (btw - many employers offer a permanent contract after 1-1.5 years if they're happy with the emplyoee, at least that's what I see in tech). Of course people in the hotel/entertainment business may have a different experience. I bet flex work is more common there.
> That actually sounds legit bad IMHO if mental issues are the becoming the norm in employment relationships.
I didn't say mental issues are the norm in the Netherlands, lol. Far from it, it's one of the happiest countries in the world according to research. I'm just saying - the law enables you to take a LONG paid leave if you have burnout. Now some employees actually get burned out, and some are probably abusing it a bit. When you're protecting employees there'll always be people who dishonestly try to take advantage.
Laws have huge margins left for judge's interpretation of what is fair and just in a given case.
The range of reasons include coming too late to work, insulting coworkers and even underperformance.
However you still have at least 1 month notice, and often more depending on the industry and seniority. Here in Zürich the standard in tech seems to be 3 month (anecdata though).
Where it gets kinda business friendly is that if you are fired for an illegal reason, it's up to you to prove that you were instead of your employer having to prove he fired you for something reasonable. The settlements are also nothing that would deter a company to behave badly I guess.
There is also a good tradition of union-like structure in the professions that need it (the ones where one is easily replaceable basically) that give workers a nice protection and benefits. It's not like in France of course but it's far from the US, based on the stories I gleaned about on HN.
>and risk-adverse VCs.
I don't know exactly how realistic this is. But from this side of the pond [Europe], it seems like, in the US there are VCs willing to throw millions [billions sometimes] at companies which are little more than a couple of guys with a silly idea, which would get you laughed out of your bank manager's office over here, if you went in asking to borrow a few hundred euros/pounds.
I've met a lot of European founders in the valley.
It doesn't happen with the one large corruption scheme taking all of the money. Its hundreds of small corruptions like a locust.
Here is a fair and neutral attempt to catalog all the schemes in play: https://www.oecd.org/gov/ethics/prevention-fraud-corruption-...
In return: Do you have an example of where the EU paid anywhere below 200% of industry standard price for anything?
I'm not arguing its more corrupt than say the US Congress. Its on par, with some weird extremes. For example: Berlusconi, a prime minister, spending public money on under age prostitutes, serving today freely as a member of the EP. That kind also owned most media companies in his country and implemented favourable regulation and tax exemptions for himself.
Here in the Netherlands, there is something referred to as the 'job carousel'. Where favourable regulation by politicians is rewarded in the future (after a political carreer) by very highly paid gigs in the industries and companies they favoured. Technically legal, and very hard to legislate against. Its a refined form of corruption, but it has all the same downsides. During Corona for example, the Netherlands had trouble getting its testing in order. Turned out we had excluded big commercial labs from consideration because we had represenatites of the 'small artisinal labs' in our taskforce lobbying. A small scale corruption, with heavy economic distortion as a result!
I'm also not perse blaming corruption on the politicians themselves. We are all together responsible for the corporate and governance culture we breed. Yet we shouldn't pretend we currently have the ability to spend/invest money effectively and fairly as a collective.
Do u have a link to that story?
https://www.ftm.nl/artikelen/coronatesten-belangen-arts-micr...
However, this stuff is paywalled by now.
https://nos.nl/nieuwsuur/artikel/2346643-testlabs-kunnen-wel...
Or, an easy video summary by Arjen Lubach (has english subtitles): https://www.youtube.com/watch?v=JUx-kdbKQys
The lobbyist, part of the Outbreak Management Team, is Ann Vossen. And she argued from her experise that there should be specific requirements that effectively kept larger laboratories from being able to be suppliers. Ironically, the trick used was the Tower of Babel (pretending language differences could be a risk)
I find it telling that it's mostly europeans who make these types of comments. The EU has tremendous barriers to labour mobility that most europeans don't even realise - differing bureaucracy and language being two of the most significant. And i cannot emphasise the bureaucratic hurdles enough when moving countries. Despite the EU crying itself hoarse about labour mobility, only around 5% of EU citizens relocate between EU countries. The majority of that 5% is people moving from poor to richer countries. The comparable number in the US is 20%.
Most rational people realise that despite all its flaws, some of the most key factors in US tech dominance are the existence of a large talent pool, all of whom speak the same language, the existence of a large market all of which operates in the same language and finally VCs willing to take on risk.
Corrupt politicians are not hobbling European tech success any more than corrupt US politicians are hobbling its success. The EU will not reach Silicon Valley's levels of success, at least not in my lifetime.
It's a chicken-egg problem, and Europe has no chicken nor eggs at the moment. But lots of grain and hay.
It has never been cheaper and easier to operate in several languages. Seriously, translation and internationalization is as cheap as it can get (and most folks in Europe are at least bilinguals!).
What they need to focus on is a unified marketplace and harmonized regulation. Now each tiny country has it's special snowflakes regulations (how to accept payment, who and how to pay taxes, huge amount of paperwork to get employees and ultra-restrictive labor laws). I mean it's good as a job program to keep EU civil servants employed, not so much to bootstrap a tech ecosystem!
The fact that the US operates entirely with the global lingua franca is a gigantic advantage that can crush competition before it even begins.
The unified marketplace is one problem but look at what has already been done - EU labour is free to move. Degrees are almost always mutually recognised. What stops further integration? It ultimately boils down to language. This is why a Greek doctor can't just move to Germany. Also why a Spanish lawyer can't easily move to Sweden. Language is a much bigger problem than people here seem to think.
These are all customer service positions (you get a prefix in front of your name but still).
That matters less with tech. You'll have a localized sales team/customer support sure, and that adds expenses, but it's not a huge burden.
Dealing with a huge amount of paperwork, taxes, special local laws of the N jurisdiction you're operating in, now that's a barrier to entry.
Law degrees might be harder to transfer generally though (easy to transfer in the Nordics at least iirc) but even in the US you need to take the bar in different states, so maybe that part isn't too different anyhow?
Actually, Might become easier in the EU in the future now that post-brexit there isn't any common law countries within the union and with mostly civil law systems it should be easier to unify principles in some sections of the law.
It's mostly northern Europe. Try France, Italy, Spain then you will change your mind.
I'm French, where I work I'm the only one on 200 people that can take a phone call in English.
Semiconductor manufacturing is really a global effort -- most device makers have fabs full of equipment mostly from US, European and Japanese suppliers. Korean equipment suppliers are also becoming more prevalent. Chinese suppliers are starting to try to gain some market share as well.
I used to work for a semiconductor equipment manufacturer a long while back so am familiar with that industry. ASML works on just lithography but there are dozens of other processes needed to manufacture semiconductors. The biggest of those in market share are the US companies Applied Materials and Lam Research. They make a quite a few of those machines. So ASML is an important part but not the only part.
https://en.wikipedia.org/wiki/Semiconductor_equipment_sales_...
FYI, a LOT of the critical components in ASML come from the US. There's no one country that can make the latest tech end-to-end relying on it's self. China seems to be trying, but it'll be inefficient.
My understanding is that Intel believed EUV would be very late. As such, it committed to Intel 10nm (TSMC 7nm) without EUV, which turned out to be too complex. At the same time, TSMC banked on EUV early. This gave them a headstart over Intel once Intel started down the EUV path.
The tables could have been reversed. EUV could have been very late, and 10nm non-EUV might have worked better. Hard to know these things when you are making plans 3-5 years ahead.
If they can execute, which I very much doubt, given the history of similar EU initiatives. The only example I can think of is Airbus. Those funds are usually funneled toward bullshit vendors that are specialised in milking such EU programs, with nothing to show for it afterwards.
I’ve met a CEO of one such company. A filthy rich scumbag that has half the government in his pocket. Ensuring that he gets the big government contracts and muscles out other companies for EU funds.
When slapped with tariffs later judged to be illegal the Trudeau government didn't do anything, allowing Airbus to get ownership of their flagship plane.
It was fortunate for Europe that the consortium had a longer-term objective as opposed to other pan-European aircraft projects like Panavia and SEPECAT.
I do agree that every continent should have a modern fab along with the supply chain to feed and repair it.
And ST Micro chips are heavenly used in IOT. Maybe it's more than the semiconductor sector and includes telecommunications, networks equipment, phone/pc CPUs.
Makes you see ST in a new light.
The thing is, micro-controllers do not need advanced nodes. They don't have a lot of logic (it's all relative, but compared to a PC CPU or smartphone SoC), so would be "pad limited" on advanced nodes: the area of a chip depends on the logic (and memory), but also on the I/O pins or pads, which take area. So if you used a too advanced node for a chip with (relatively) little logic and still enough I/O pads, some die area would be wasted as the minimum area would come from the pads. It's a situation you want to avoid, as there's no point paying for a more costly finer nodes (with more leakage) to waste it by not using the silicon area.
Another factor is that micro-controllers also embed Flash on their die, to have a very integrated solution with logic, RAM and Flash storage on a single die. This puts a limit on the node density. Flash is OK down to 40nm. Then from then one must move to more advanced, and still young and costly, alternatives (MRAM).
So if your business is driven by micro-controllers, there's no real point in pushing toward node density ("low" nodes). ST does process optimization as everyone, but at "medium" nodes best suited for µC. There's no business drive for them to chase TSMC.
It's similar for other Eurpeans silicon makers: they tend to be on specific segments where going for "low" nodes like 5nm simply doesn't make sense. The relatively advanced fabs in Europe are from Intel and Global Foundries (both US). And GloFo stopped the race to advanced node at 14nm, as it was too expensive.
Why MCU-bound application logic are simpler than PC apps? If that is the case, we shouldn't have seen improvement in embedded MCUs in 20 years, yet every year the clocks are increasing, the flash/RAM size is doubling, ... I have a theory/fear that in future, the line between embedded-dev and web-dev would be blurred...
As for low volume designs, if reliability is important, rust would be a good candidate.
The point is if China decide to do this, there isn't anyone who can stop them.
Taiwan's history is complicated. it have its own military. its people pay taxes to Taiwan, not China. Taiwan's safety is its own responsibility. its more than possible that China can take Taiwan by force but at what price. its not like Taiwan's military have no teeth.
After it became part of China again, it is not possible to interviene militarily in Beijing - HK relationship - that would be an invasion of China and violation of international law. You could not phisically station an army in HK.
Even if China has violated an agreement with UK regarding HK autonomy, thay does not give you legal grounds to start a war.
In cotrast, you could deploy the entire US military in Taiwan tomorrow, it would be legal and peacefull (provided Taiwanese Gov. Invited it). Additionally, invading Taiwan is actually extremely difficult because of terrain, just because you have a massive land army does not mean they can just teleport to Taiwan. It would be basically a D-day style effort against a well entrenched opponent.
So it is unlikely we will see phisical takeover of Taiwan in the near future, as so far we have no indication that China is looming for a causualty heavy war with little gain.
This is true, but I view China's economic growth more as 'reversion to the mean' when their population is 3x that of the US.
See Cuba and the Soviet Union
Is their military power proven and able to assert power or enforce its will?
Will America maintain its might?
Lots of better stuff in the US, until recently when the stagnation in the US made it more feasible (and politics made things scarier).
It ain't much, but it's honest work.
far from it