Could you make the same argument for anything: "Why would anyone manufacture chip lithography equipment in Europe?" Without knowing that most of that equipment is made there, I would have asked this question. This is about mostly automated foundries and already expensive high skill low number labor.
The largest manufacturer of semiconductor manufacturer equipment is Dutch, so it doesn’t seem that outlandish.
More generally, the current market is unhealthy. There are only two companies fabbing high end chips (not counting Intel, who generally only fab their own stuff), and even then only if you’re generous to Samsung. In the long run, that’s not good for anyone, and it’s an obvious analogy to Airbus.
The entire chapter is interesting, but it also mentions how TSMC came to be:
> The government came to realize that it did not yet have an industry—only one company. The private sector was still fearful of risk and could not raise enough money to start new firms. Once again the government put up money, this time to start TSMC. And this time it applied a condition: The government’s share would be less than 50 percent. Lacking enough funds from the private sector, the government made an offer to Philips, which put in almost 35 percent of the initial investment; this pushed the private sector’s share above 50 percent, and TSMC was born. Again, the technology and all the people came from ERSO, including about 130 engineers, the 2-micron CMOS developed at ERSO, and its latest 6-inch-wafer fab. TSMC had no product, only the fab technology, which matched well with the idea of a pure-play foundry.
I don't get why Europe is so bad at this kind of thing.
Because Taiwan is a single, small country, with a single government, and a great incentive to succeed in semiconductors.
The EU is a conglomerate of 20+ hyperbureaucratic states where the incentive is to drag these initiatives for years and billions, get nowhere, and repeat. A lot of people will make a lot of money, regardless the outcome.
I'd put it slightly differently. Germans won't fund Philips. France wants CI Honeywell Bull, not Olivetti.
If you want to invest, you either back new starts which wind up being a race to host them, or you back collaborative projects which force Bull, Olivetti, ICL, to work together. Sorry, they just never really work. Too much competitive tension.
Taiwan, its kind-of a bit more apparent. The nasty neighbour is just across the straights. You better get your act together.
This is also why there won't ever be a European Silicon Valley. Paris isn't going to support any Europe-wide effort to create one that isn't located in the Hexagon, nor will Berlin support one that isn't somewhere in Germany. London pre-Brexit would not and did not support anything that might take away from Silicon Fen.
While other US states encourage their own tech centers' growth, Floridians and New Yorkers and Texans are all aware of, benefit from, and are proud of Silicon Valley being in the US.
This is a pretty good explanation for why most EU initiatives can't work. It's because they have to make everyone happy, which is an unattainable goal.
Arm can be seen as a direct product of the UK's Computer Literacy Project, and slightly less directly the broad base of games developers in the UK could be considered another output.
The Transputer was less of a success, but left a legacy that I think can be traced through to Graphcore today.
At a certain scale, things don't come out of thin air.
There's almost no chance of something like this happening on it's own, although the initiative is probably doomed, it could feasibly work if done right.
Europe has gigantic research funding in this space, and arguably more fundamental research than probably even the US (which often focuses on gettting things to msrket). Problem is that most successful EU companies get bought out by US (and previously also Chinese) companies before they scale enough.
This. ETH Zurich, and the IBM Zurich relationship comes to mind. And, The various institutes of like nature. People congregate there to do smart stuff, but then fly off to the land of IPO and giant cars.
People forget how much modern tech comes from Eindhoven. It just wound up being made in Asia, it was invented in labs at Philips, in Eindhoven. Or, the amazing tech in the oil industry which stems in part, from the work done for Radar by GEC Plessy Marconi. Radar for bombs and missiles is in the end, easy to repurpose for Geo. And.. North sea oil happened.
Europe makes amazing hugh value widgets like the machines TSMC use to make the chips Europeans buy. The key point here is that Europe is pretty crap at buying and operating these machines themselves. At scale. TO make widgets people want to buy. OUTSOURCE! (and.. all the things which stem from it)
> But he acknowledged at the report's launch that it was still unclear whether this would involve offering tax cuts to woo TSMC, Samsung or another major player to establish a new site within Europe.
Sounds like it'll be another half-assed EU project that won't amount to anything meaningful. They should be looking at home-growing talent (research labs, universities, programs) and rebuilding the European industry (joint partnerships, funding, subisidies, etc.), instead of relying on overseas companies for such a crucial and strategic issue.
The link I see from the few words GP used, one of the core underlying messages behind the "Make America Great Again" campaign was to bring back manufacturing within America.
Subsidizing and growing local production of computer chips would be quite similar, so I can see a parallel there.
GP really should have expounded on their view though, rather than leaving the burden to interpret their words on us, the readers.
EDIT: The derision you may feel might be because many shared the view that bringing back manufacturing was pointless and that bringing the capability back did not mean bringing back the jobs (automation). There was also the widespread conflation of Trump, his other policies, and the MAGA slogan... so the slogan is probably not as useful now.
The problem is how to do that. Samsung is a European company in a sense as they have a European legal entity, HQ, ... Similar for any other big company.
So how do you even design a project that excludes them, includes the companies of the 27 EU countries, and doesn't fall foul of WTO rules and trade deal provisions? If you deal with a single country you can get quite creative, but as EU they have a hard time being exclusive while keeping the veneer of inclusivity as e.g. UK, China, etc like to do.
Samsung is a global Korean conglomerate. In your sense, Samsung is an Indian company as well because they have offices and have a business registration as Samsung Pvt Limited.
[1] This article talks about electronic components in general, not just chips, plus "work together to bolster Europe’s electronics and embedded systems value chain".
[2] is about quantum computing
Basically EU is boosting research with money set aside for this purpose, hopefully being spent on research labs, universities, programs.
I've had a number of conversations recently on how semi conductors could be the new oil in relation to geo politics and it seems that many policies are coming out that reinforce that.
It'll be interesting in the coming decades whether or not we can navigate this one peacefully or not. There is likely hope though as it's not really limited by natural resources or geography, at least right now, but rather on intellectual property which is inherently infinite except for the artificial constraints we've applied to it.
The article says it costs $20B to build a state of the art fab. The USA is just about to pass its second $2T stimulus package. I’m betting a fab is not in there.
The EU doesn't have the talent or the resources, especially not if they're planning on fleecing engineers like they do with software engineers (which earn 1/3rd or less the equivalent position's salary in the US).
There's no reason to be in the EU as an engineer, unless you like making trade-offs that leave you in a net negative position (no, free healthcare is not worth >$100,000 less salary).
High taxes and handouts to corrupt companies make it too hard to compete inside the EU. Everything is about winning the EU grants, and the better you are at winning EU grants, the more profitable your company gets.
I prefer to be happier with live, a better quality of living than earning $100.000 more in salary. Just like in Europe in the US there are only specific areas were you would earn such high salaries.
Don't think you would earn that much in Holland, Michigan but happy to be corrected :)
What you are saying doesn't make sense. If engineers can earn $100k more then why wouldn't private companies build the fabs in the EU?
You can't conjure money out of nothing.
If anything, cheap engineers would make it easier to build the fabs and stay competitive internationally because that is how Taiwan is competing.
Europe has something arguably more valuable than semiconductor manufacturing - they have ASML Holding, which produces the equipment that makes the semiconductor manufacturing possible. In the gold rush, they're the ones selling picks and shovels.
Yes, but the EUV license that underpins ASML's technology is owned by the EUV LLC, composed of Intel, AMD, and the US Department of energy, all those other members being US institutions, only ASML being a non american organization.
Yes, but you are forgetting a massive amount of the research for EUV was done at and 'owned' by Euclides, ASML's EU based R&D center. It is more of a cross-licensing thing than the US companies being the sole owner of the tech.
37 comments
[ 3.3 ms ] story [ 85.2 ms ] threadMore generally, the current market is unhealthy. There are only two companies fabbing high end chips (not counting Intel, who generally only fab their own stuff), and even then only if you’re generous to Samsung. In the long run, that’s not good for anyone, and it’s an obvious analogy to Airbus.
https://news.ycombinator.com/newsguidelines.html
Anyone else worked through 5th generation, the Alvey report and the MITI catchup?
What does work is sustained organic investment in fundamental science and an environment which stimulates SME.
Its the transition to world scale which seems to defeat European initiated planning.
https://www.nap.edu/read/10677/chapter/9
The entire chapter is interesting, but it also mentions how TSMC came to be:
> The government came to realize that it did not yet have an industry—only one company. The private sector was still fearful of risk and could not raise enough money to start new firms. Once again the government put up money, this time to start TSMC. And this time it applied a condition: The government’s share would be less than 50 percent. Lacking enough funds from the private sector, the government made an offer to Philips, which put in almost 35 percent of the initial investment; this pushed the private sector’s share above 50 percent, and TSMC was born. Again, the technology and all the people came from ERSO, including about 130 engineers, the 2-micron CMOS developed at ERSO, and its latest 6-inch-wafer fab. TSMC had no product, only the fab technology, which matched well with the idea of a pure-play foundry.
I don't get why Europe is so bad at this kind of thing.
The EU is a conglomerate of 20+ hyperbureaucratic states where the incentive is to drag these initiatives for years and billions, get nowhere, and repeat. A lot of people will make a lot of money, regardless the outcome.
If you want to invest, you either back new starts which wind up being a race to host them, or you back collaborative projects which force Bull, Olivetti, ICL, to work together. Sorry, they just never really work. Too much competitive tension.
Taiwan, its kind-of a bit more apparent. The nasty neighbour is just across the straights. You better get your act together.
While other US states encourage their own tech centers' growth, Floridians and New Yorkers and Texans are all aware of, benefit from, and are proud of Silicon Valley being in the US.
The Transputer was less of a success, but left a legacy that I think can be traced through to Graphcore today.
I suppose if they tried to replicate that sort of pump priming failure they would be thought of wasting money.
That's the part Europe seems to miss: making endogenous industry in territory
At a certain scale, things don't come out of thin air.
There's almost no chance of something like this happening on it's own, although the initiative is probably doomed, it could feasibly work if done right.
I'm doubtful but there is a path that could work.
People forget how much modern tech comes from Eindhoven. It just wound up being made in Asia, it was invented in labs at Philips, in Eindhoven. Or, the amazing tech in the oil industry which stems in part, from the work done for Radar by GEC Plessy Marconi. Radar for bombs and missiles is in the end, easy to repurpose for Geo. And.. North sea oil happened.
Europe makes amazing hugh value widgets like the machines TSMC use to make the chips Europeans buy. The key point here is that Europe is pretty crap at buying and operating these machines themselves. At scale. TO make widgets people want to buy. OUTSOURCE! (and.. all the things which stem from it)
Sounds like it'll be another half-assed EU project that won't amount to anything meaningful. They should be looking at home-growing talent (research labs, universities, programs) and rebuilding the European industry (joint partnerships, funding, subisidies, etc.), instead of relying on overseas companies for such a crucial and strategic issue.
Subsidizing and growing local production of computer chips would be quite similar, so I can see a parallel there.
GP really should have expounded on their view though, rather than leaving the burden to interpret their words on us, the readers.
EDIT: The derision you may feel might be because many shared the view that bringing back manufacturing was pointless and that bringing the capability back did not mean bringing back the jobs (automation). There was also the widespread conflation of Trump, his other policies, and the MAGA slogan... so the slogan is probably not as useful now.
So how do you even design a project that excludes them, includes the companies of the 27 EU countries, and doesn't fall foul of WTO rules and trade deal provisions? If you deal with a single country you can get quite creative, but as EU they have a hard time being exclusive while keeping the veneer of inclusivity as e.g. UK, China, etc like to do.
[1] This article talks about electronic components in general, not just chips, plus "work together to bolster Europe’s electronics and embedded systems value chain".
[2] is about quantum computing
Basically EU is boosting research with money set aside for this purpose, hopefully being spent on research labs, universities, programs.
[1] https://www.reuters.com/article/eu-tech-semiconductor-idUSKB...
[2] https://ec.europa.eu/digital-single-market/en/news/nine-more...
It'll be interesting in the coming decades whether or not we can navigate this one peacefully or not. There is likely hope though as it's not really limited by natural resources or geography, at least right now, but rather on intellectual property which is inherently infinite except for the artificial constraints we've applied to it.
There's no reason to be in the EU as an engineer, unless you like making trade-offs that leave you in a net negative position (no, free healthcare is not worth >$100,000 less salary).
Don't think you would earn that much in Holland, Michigan but happy to be corrected :)
If anything, cheap engineers would make it easier to build the fabs and stay competitive internationally because that is how Taiwan is competing.