To think that other countries should also be able to make semi-conductors, not just Taiwan

[MargoLivebetter]

I know very little about technology manufacturing, but I have to wonder how we have got into a situation where Taiwan produces 90% of the world's semi-conductors? How has this happened? Why aren't other countries pulling their industrial and technological fingers out and making some of their own?

I am a bit of a tech dumbo, so please explain to me as you would to a primary school child.

@MargoLivebetter I don’t pretend to be an expert

But I don’t think this is something that we can easily get more into, we have no baseline for it and no infrastructure and no staff

Also bear in mind, Taiwan having to import most of the raw materials from Japan, aren’t they? And there was a pinch point at the beginning of all this when Japan was a bigger player than Taiwan.

to some extent it’s just that different countries will choose to be experts in different ways isn’t it? It wouldn’t have made sense for us to make a big industry out of something where most of the raw materials are in Japan I guess.

Like I say, not an expert.

Think strategy rather than policy.

The strategy is delivered by non departmental public bodies. As I said, established under the haldane principle which dictates that R&D investment is decided by research not government.

@EmeraldRoulette thank you. It is interesting to get insights on this - although just having the temerity to ask appears to have made me a populist in some posters opinions!

Thank you @DeathBanana I can see that the UK Semiconductor Advisory Panel last met in December 2025. This was the summary of the meeting's discussion:

• The Panel discussed skills and the need to raise awareness of the industry to develop the talent pipeline across both design, manufacturing and academia.
• The Panel discussed the reforms to the London Stock Exchange, and the private investment landscape in the UK.
• The Panel discussed how the defence industry relies on the semiconductor sector and opportunities to develop more resilient supply chains.
• The Panel heard about the British Industrial Competitiveness Scheme (BICS) consultation, and opportunities to feed into the development.
• The Panel discussed the opportunities represented by growth from the AI sector, but emphasised that to benefit from growth would require an appropriate skills base.

Where do you suggest I go from here to understand more about delivery?

£1000 pa? With inflation it will be many times that before we even see a finished product.

If it's left to the private sector they want a ROI, so price will be high as they won't accept breaking even in ten years time.
If it's public sector investment, what gets cut to fund semi-conductor production.
Are we manufacting only for domestic consumption unlikely or are we planning to foolishly compete against Taiwan.
We can't magic up the components we don't already/produce, and what we do with the toxic waste etc. (See current protests against data centres)

There's usually complex and good reasons why some countries specialise in manufacturing some products.

This sounds like it’s at the beginning of making something happen? Better than nothing but seems a long way to go.

@EasternStandard yes, it sounds like the beginning to me too, but @DeathBanana and @Singasonga seem to think I am missing what is actually being delivered and I'm trying to find out more! I want to think that there is actually way more going on than I'm currently aware of.

@trumpredcard I don't really know yet what domestic manufacture currently looks like, so I can't answer your questions! I'd like to think that we would have some resilience if for any reason semi-conductors ceased to be fothcoming from Taiwan, but truthfully I don't know what that would look like. I'm trying to better educate myself and find out more.

The main thing is that Trump made all of this common knowledge last year, which is again 10 years after the fact, because he suddenly realised they are reliant on Asia. Populists don't like thinking too hard and solutions with problems that can't be papered over, like geography when they are trying to tell their population they can have it all and fuck everyone else.

Taiwan has managed to stay out of China's grasp largely because it had the foresight to become globally integral to progress. Like China it thought ahead and invested in its industry. No one wants China to take over this stronghold.

Another example is with recycling, we dump everything on Africa rather than actually solving the problem and potentially coming up with a way to become completely integral to the world. We had so many projects about enzymes eating plastics but no one up until a couple of years ago realised how much gold, for example, was being shipped over and dumped Africa in tech within chips. It really seems ridiculous that we only realised this years after using Africa as a tech bin and are still only just turning it around with minute steps (IIRC the Treasury is doing something WRT recycling precious metals now) on a rather small scale.

You left out skills.

Google “what is UKRI doing to implement the governments semi conductor strategy”

Ok, so I have found the 2025 budget allocations from the Department for Business, Innovation, Science and Trade for UK Research and Innovation until 2030. Within the R&D investment of £86 billion, it appears that there is a sector described as "Digital and Technologies: Advanced Connectivity Technologies (ACT), Semiconductors, and Cybersecurity"

And within that sector the following allocations have been made from the overall budget:
Strategic government and societal priorities £202m
Supporting innovative companies £516
Combined investment £718

Does that sound like enough?

We had a lot more before Brexit. Populists again decide collaboration is not needed and we can fund everything we need independently. See also farmers and food sustainability minus EU subsidies.
Thanks Russia! Although turkeys and Christmas spring to mind with a lot of farmers and their signs in fields...

Pressure and dire need helps with innovation. Good life and globalisation makes it easier to buy things instead of manufacturing everything at home.

Ukraine now has the most high-tech army in Europe in some practical aspects - because they have to, because anything they get gets tested and improved immediately.

Taiwan uses their technology as a kind of safety measure, hoping the word's interest in their semi-conductors will make the word care about them and that this interest will deter China. It has worked so far.

Other countries do produce semi-conductors, but typically can't do the latest generation.

I don’t think even with that it would be realistic to replace the role of Taiwan

Really to me, the important stuff is we need semiconductors for infrastructure so gas, electricity, water… and then of course food production.

(I have no idea why anybody’s talking about populism either by the way).

Populists have shaped our responses to these things, as I said an example is Brexit, where quick fixes win over actual thoughtful processes and progress.

We have issues with water because AI is going to use data centres, which will affect water supply in areas already struggling with droughts. Add to that the thousands of houses still being built on these areas and we have potentially a deadly situation at a time when our climate is changing rapidly. Again, zero foresight despite decades of warning. Speed seems to trump logic.

It's rather funny (or worrying) that this discussion is taking place in the nation that gave the world Adam Smith and the seminal "Wealth of Nations", with it's case study of how pin manufacturers used the division of labour to maximise profits efficiency.

The is no technical reason why the UK (for example) could not become a massive chip manufacturer to the world. It would cost. And be unlikely to make a profit. But it could be done.

I never cease to be annoyed when people try to pretend they haven't made a decision because of some spurious guff meant to silence the less critical thinkers.

We could put a person on the Moon if we wanted. Nothing impossible about it at all. So the reason we don't, is because we choose not to. It's a choice not a law of nature.

The reason that matters is because there are so many issues around us (in the world as well as the UK) which are because of choices, not because of any fundamental structure of the universe.

I looked it up and got this

By focusing on intellectual property (like 's architecture) rather than physical factories, the UK requires significantly less capital to generate high gross value added (GVA).

It effectively secures leadership in specialized fields like silicon photonics, flexible chips, and power electronics, rather than trying to beat Taiwan or the US at traditional computer chips.

@MargoLivebetter ,

I would thoroughly recommend a book called ‘Chip War’ by Chris Miller. It is a really remarkable field. This was written before AI got big, so I imagine it is really important now.

Chips are made by etching lines into silicon using lasers. The most advanced CPUs now have about 180 billion (180,000,000,000) transistors on them. This means that each transitor has a width of 2-3nm (1nm is equal to a billionth of a meter). To put this in scale, visible light has a wavelength of 450-700nm approximately! So, the etching is done with extreme ultraviolet lasers.

The machines that do this etching are made by ASML, in The Netherlands, which is a monopoly. Each machine costs up to $400 million and comes with its own team of technicians. Any contamination in the air will make the process fail, so the air has to be kept at an incredible level of purity.

Hopefully, you can see now how hard it is and the barriers to entry of building new chip fabs. The cost is around $20 billion and the timescale is 3 to 5 years. In addition you have to get the machines from ASML and have a skilled workforce to use them. I think the U.S is doing its best to slowly create alternatives to Taiwan but it is clearly proving difficult.

Thank you @Newbutoldfather that sounds like a really useful book to look at.

Despite all my efforts and being sent off to Google things (!), I'm still finding actual examples of delivery to be very limited. And a Govt budget of £516m to support innovative companies in a broad sector that includes semi-conductors seems fairly low to me.

This tiny island had had its time as an industrial power - built on access to natural resources and/or the ability to exploit others. We had first mover advantage too. These advantages have now passed to others as the global economic structure has changed. We need to innovate rather than attempt to replicate.

telling you to google things is not (!) it’s the way you will learn. You’ve moved from why is no one doing anything to £516m doesn’t sound much. That progress right. In the course of one interaction on a parenting forum.

no £516m is not much. Successive governments have kept R&d funding flat (while all things continue to increase) but it’s not nothing.

I worked a Lam Research in California, so I can offer insight. Setting up a FAB (fabrication units and clean rooms) costs approx $100,000,000. The technology they use on a silicon chip is nanotechnology it’s like a city built upon a single chip, so the build and qualification period to get the fab up and running is close to 10years. This means the barriers to set up and enter the market are very high. Due to tensions with China/Taiwan, the USA agreed with TSMC that they would build a fab in the US, this is probably 5+ years into the build. USA learnt from the last semiconductor shortage that the supply chain is fragile so they took action. UK has done nothing.

You also have to mechanically insulate the building so that vibrations don't interfere with the process. Which rather limits where you can site the fab.

If we could crack cooling, putting the buggers in orbit is an idea.

I suspect the Moon is a little too shaky ... but that is another tack. After all if you can build an automated factory on earth .....

Forgive me @DeathBanana but I find your post deeply patronising. I'm trying to have a discussion.

Whilst I may now have learnt what the Govt budget is for supporting innovative companies in a sector that includes semi-conductors, I remain none the wiser as to actual production itself - of if that is even remotely feasible.

@WizdomE wow, thank you. 10 years!!!!! Bloody hell.

But how practical is it for us to try and do this? You’re talking about actually setting up as an industry that would be a mirror of Taiwan - am I right?

How much money do you think we should invest? To some extent I think we’ve got to accept that if there’s been a monopoly for decades, we can’t now step up and compete with that monopoly.

@EmeraldRoulette judging from what @WizdomE has said, it doesn't sound as though it will be remotely feasible for us. That sounds like a significant vulnerability to me.