No, Europe is not a factor in cutting-edge chip assembly. That is a tiny fraction of the overall supply chain to manufacture a finished, compute-ready semiconductor. There is sense in not trying to compete with the "sexy" part of chip manufacturing, and instead concentrating on the steps of the supply chain you are already good at.

I dont think there is any hard fact behind your statement. EU has a number of semiconductor companies and fabs, though they tend to focus on embedded/industrial applications. Some went bankrupt (e.g. a big German RAM manufacturer) and the number dwindled, but it's not like there's nothing.

Like in anything high tech you need either massive subsidies or network effects, this is why things have built up in Taiwan, Shenzhen and South Korea, with remnants in the US, EU and a bit more Japan as all three thought they can let the market play out, fell a bit behind, and then the network effects won out.

Similarly the car industry - Germany (and Japan) rose due to massive investment and network effects, then the shareholders put in CEOs with a clear mission just to milk the cows and get wealth out of the system; this gave China a chance to go all in on investments and overtake them on innovation and price. This has nothing to do with regulation, just with short-termism and the inherent diseases of a "pure" capitalism.

That there are few new innovative companies coming up is probably more due to regulations (need to figure out how to make the product work and be legal in various regimes). In turn that they don't scale is due to less financing options and a much more fragmented market (to sell in France and Spain and Poland you need to translate everything and adapt it to local expectations and payment standards.. to sell in Florida, Minnesota and Texas you just need to throw a single English version up.

A shame that the national governments keep blocking any real integration of the financial and regulatory regimes at EU level, this would solve the startup and scale up issues to a large degree. The EU just proposed "EU Inc" as a solution, not sure this will not be ground up in 27 regimes...

They're making the point that you're wrong. The CHIPS act under Biden is a big subsidy to bring modern fabs to the US and without it there wouldn't be any orders in the US either (which has less regulation). The EU came up with a competitor to that subsidy but it wasn't generous enough to compete (neither regulation nor bureaucracy at play here).

On a side note, none of the TSMC fabs in the US are cutting edge. They're all at least a generation behind.

Isn’t ASML a European company?

> primarily due to regulations and bureaucracy

Or, more likely, due to energy prices.

It's rhetorical because I see that complaint about the EU too often and I just want to point out what you have said but in my case I'm ridiculing the moralising lecture about the dire consequences of anti-free-market behaviour.

70 machines does not mean 70 orders

And about 300-400 DUV machines a year

Hopefully for RAM and RISC-V ultra-performant µarchitectures.

Yeah, I am advocating for I would like to happen (hopefully INTEL and AMD will put a RISC-V decodecs on their µarchitecture out of good will... yeah since they hold the x86 and x86_64 IP then excluding all alternatives de-facto... the incentive is.. negative.

If I am not mistaken, EU is currently pouring money in defence. For defence needs, one state-of-art chip build line is enough... the EU armies could even share it.

Large shopping malls would fail the air purity requirements for semiconductor manufacturing.

EUV lithography and etching rooms require ISO class 1 clean rooms. Those can have a maximum of 10 particles of size 0.1 microns per cubic meter.

While a shopping mall can get away with 2 - 4 air changes per hour, an ISO class 1 clean room requires 400-600 an hour, or about 10 a minute. Airflow needs to also be controlled: vertical and laminar. Every human entering the cleanroom needs to undergo an air shower with speeds of 60-90 mph to shear off loose skin flakes, hair and lint. This area needs to maintain constant positive air pressure to repel outside dust as humans move in and out.

In the EUV lithography systems, temperature control is measured in milliKelvins (0.001 C). If a wafer stage has a temperature variation of 0.05 C, the silicon expands by a few nanometers and the multimillion dollar batch is toast. Meanwhile in the same room, the EUV machine is generating extreme ultraviolet light using a plasma light source that gets to above 200,000 C. That's 40x hotter than the surface of the sun.

> I always thought of these as high-precision, not-too-high-volume processes

EUV rooms are churning out about 150 wafers an hour, 24/7. That sounds like not too high of a volume, but each wafer is about 500 cpus. So 75,000 cpus an hour.

TSMC's Arizona fab needs 200MW of electricity.

But even more importantly, they need 2 separate completely independent substation feeds, since a 15 second interruption can cause 0.1B Euro of losses.

Finally, and here is where Europe really really suffers, generally you need these things to be planned, approved and built within a year.

Check out asianometry on YouTube for deep dives into many parts of semiconductor manufacturing, including the lithography processes ASML uses.

They have hundreds of machines that use microwaves, turbo molecular pumps like cvd, ion implantation, plasma etching, idk what else? I'd say it's just the sheer density of equipment within. An average fab has 2000 tools aka machines for wafers alone

Most fabs build their own separate power grid.

And Japan, South Korea, and Taiwan are also energy poor states that have been able to maintain capacity.

The energy excuse is true for some manufacturing processes like automotive in CEE which were pants that were built in the 1990s to 2000s and leveraged Russian ONG, but is a red herring for a lot of other manufacturing processes in Europe.

European states have a coordination problem as Draghi pointed out 2 years ago.

AFAIK the problem in NL is due to chronic underinvestment in the grid infrastructure since the big privatisation wave. Combined with the heavy subsidy of industry causing the issues.

Solvable but we need to be much smarter and redirect the fossil subsidies to reducing risk for grid-scale batteries and adding more grid capacity.

A consequence of a totally self-imposed handicap.

We can’t let industry be built, because the power infrastructure is so shabby, because the state can’t afford to improve it, because the economy is in tatters, because there is no industry, because we can’t let industry be built

How? I thought there was still tons of demand for every node that has launched in the last 15+ years.

For better or worse, Ireland’s extremely lax tax rules resulted in a ton of multinational corporate cash ‘stuck’ there with companies having the option of spending there roughly tax free or repatriating back to the US at some rate higher than that. It also really helps sell the ‘of course an Irish subsidiary owns all of our IP’ tax claims if there is actual spending there, hence you got all the pharma investment and the capital projects from others like Intel and Apple.

>construction kicked off in 2019 on a €17 billion new fab development which will double

I think what the OP wanted to highlight is right here: in China or Taiwan etc. this would not have taken 7 years.

That's just low taxes pushing against everything else in the Irish regulation.

India is more than a little behind China. Which is depressing, considering that the two countries were equal in GDP per capita as recently as 1990: https://m.thewire.in/article/economy/from-1960-to-1990-india...

> a little behind China on the same trajectory

In my opinion, India is around 30 years behind China. This is unlikely to change for several reasons. However, the primary reason is that no Indian government ever had a manufacturing policy that comes close to China's. The service economy is strong but employs a tiny percentage of the population.

Most of Indian government spending is on welfare followed by endemic corruption. It leaves very little money for nation building.

> another billion people with a clear enough path

The opposite, all historical clear paths i.e. manufacturing/export led model is closed/closing (geopolitics + being automated away). Less clear service export also closing. The growth path that employs 100s of millions and 10s of millions are not available to Indian.

Reality is even PRC with most favorable conditions still left behind ~400m people because there's only so many opportunities. It's one thing to have ~1B who made it supporting ~400m who didn't vs India trending towards 1.7 B... and if they end up with 500m that makes it, and 1.2B that doesn't, then things may get very ugly in terms of governance. Now that 500m middle/upper class is large enough that India will in aggregate "make it" as geopolitical power, as in climb the tech tree, but due to sheer population and opportunities that looks now foreclosed, most of Indian's statistically, simply, won't. And when numerator/denominator of haves/have not too skewed, there's nothing haves can do (redistribute) even if they wanted to, which lbh, they won't. IMO this pretty much baked in dynamic for India for next few generations, unless we discover abundance.

> It's is heartening looking at a country like India and thinking that it only seems to be a little behind China on the same trajectory.

Depends on what you mean by "a little behind". Reading Patrick McGee's "Apple in China" (https://en.wikipedia.org/wiki/Apple_in_China) made me think otherwise.

> a little behind China on the same trajectory

I'm hoping for India development but that is quite far from the truth yet.

Every material markers indicate the EU economy is in contraction since about 2007. It's the degrowth boogeyman nobody wants to believe in but everyone will get... Cheap and unlimited fossils are not coming back and without them gdp growth is just a mathematical artifact

not to mention that ASML is likely majority owned by Americans

a.) Lack of doesn't imply inability b.) There are plenty of chip fabs in Europe

that's not even remotely true, just because you haven't heard of companies doesn't mean they don't exist. Though it does make sense because tech in Europe is a lot more focused on B2B than B2C, but the latter should definitely be funded more.

Zeiss?

Dunno, honestly, most HN people have an extremely narrow view of what constitutes "tech" (= web commerce and adware) and wouldn't survive 5 minutes in an actual engineering company. Europe is all over the supply chain of just about everything that is hard to make.