News: 1755689649

  ARM Give a man a fire and he's warm for a day, but set fire to him and he's warm for the rest of his life (Terry Pratchett, Jingo)

Fried chips: UK's nascent semi industry risks faltering

(2025/08/20)


Feature It's not easy to grow a national chip industry. Semiconductor startups are a risky investment. They chew through early-stage capital, often with little to show for it, making them a long-term proposition. Those that pay off can deliver big, but success is far from guaranteed.

Grow a new Arm: UK advisory body wants investment in local AI chips [1]READ MORE

Yet at the same time, processors are becoming more strategically important to governments, to the point where governments are banning shipments to certain countries and [2]putting trackers in chip shipments to see where they end up.

So many developed countries are looking at how they can play in this market, both for economic and national security reasons. That's why in May 2023 the UK government put a national semiconductor strategy in place as a plan to elevate its position in this business.

In fact, just [3]yesterday , a report from the Council for Science and Technology noted that the UK has a "niche and highly specialised chip industry" that is never going to be able to compete head-to-head with global big hitters like the US, Taiwan, and Korea, meaning the country needs to channel investment into areas where it can lead.

The UK has always been more of a designer than a builder of chips. It certainly has some capabilities, with around 25 semiconductor manufacturing sites there. However, the country doesn't have the kinds of world-class fabrication plants that we see already in places like Taiwan and now the US that churn out huge numbers of general purpose processors with tiny die sizes.

[4]

Instead, it focuses on more specialized designs. Since 2006, the UK government has been investing in compound semiconductor technology through the Engineering, and Physical Sciences Research Council (EPSRC). This is a materials-focused solution business that uses different kinds of chemistry to produce chips with specific properties, such as optic sensitivity and power management. Compound semiconductors make their way into applications ranging from LIDAR components in autonomous vehicles through to quantum computing.

Growing the domestic industry

The strategy had three main goals: growing the domestic sector, strengthening the supply chain, and bolstering national security.

The focus on growing the domestic chip industry played to the UK's design strengths. The government, aware that playing catchup with East Asia would be difficult, decided to lean into design and R&D, serving its base of over 100 existing design companies in the UK.

[5]

[6]

While design and IP is the strategy's primary focus, and we're not likely to see any legislation to attract huge modern TSMC-style fabs, that doesn't mean manufacturing is off the table. The R&D commitment goes beyond chip design into semiconductor fabrication (both wafer production and post-lithography packaging), along with system architecture (the building of systems that use these chips). The government will also allow eligible UK companies to apply for up to £1.5m to help scale up manufacturing capacity. That's a drop in the water when it comes to the semiconductor sector's high manufacturing costs, though.

The government's approach resonates with industry leaders like Jutta Meier, recently appointed CEO of compound semiconductor fabrication company IQE. This Cardiff-based company, which has continued to build more capabilities at its Newport foundry, has emerged from a cluster of compound semiconductor companies in South Wales to extend its reach from the US to Taiwan.

[7]

"The UK's competitive edge lies in compound semiconductors, design, and photonics, not chasing silicon megafabs," says Meier. "South Wales shows what a focused cluster can deliver."

There will also be tax breaks, although like the rest of the strategy's initiatives, these are designed to help build capabilities from the bottom up, rather than to attract heavy-hitting semiconductor whales from elsewhere. Small- to medium-sized enterprises devoting at least 40% of their expenditure to R&D could claim a 14.5% tax credit.

Part of the domestic growth plan includes shoring up a skills base that Whitehall has admitted is sorely lacking. The government instigated several initiatives designed to stoke the pipeline of new talent from high school through to university by increasing investment in STEM learning. It also committed to investments in industry-led learning to get grads up to speed in specialist real-world chip design jobs. That capability gap between grads and productive workers is common in the tech sector, and has also been a traditional issue in areas ranging from software development to cybersecurity.

[8]

Ultimately, graduates tend to go where the money is. So the government has promised to pony up a considerable amount. It committed up to £200m in investments by 2025, and up to £1bn in the next decade.

It has also promised a UK Semiconductor Infrastructure Initiative that would help get startups and others access to the infrastructure they need to deliver on this vision, and a new UK Semiconductor Advisory Panel to help execute on all these commitments.

Supply chain and national security

The government's supply chain strengthening plans under the strategy included the publication of industry guidance, and the bolstering of relationships with external suppliers and like-minded governments to create stronger supply networks. Contingency plans were also on the table for handling supply-chain emergencies. And the government promised to focus on plurilateral relationships (that is, small multi-country alliances) to keep partnerships strong.

National security focuses on two key areas: ownership, and cybersecurity. The strategy acknowledged that investment in competitively sensitive industries like semiconductors was fraught with potential problems, and that monitoring who attempted to take a stake was a good idea. It promised to use the National Security and Investment Act to do this.

The government had already done this before, spiking the acquisition of the UK's largest semiconductor fabrication facility, the Newport Wafer Fab, by China-owned Nexperia under that Act in 2022.

On the cybersecurity front, the strategy focused on securing its hardware to ensure that the design process was more resilient against attack, and would also concentrate on building more security directly into chip designs. Hopefully this will help to prevent debacles like the SPECTRE/MELTDOWN side-channel vulnerabilities in the x86 architecture.

Performance to date

That's the vision. But how well is the UK delivering? Commentary ranges from the damning to the diplomatic, depending on the source. Roy Illsey, chief analyst at market research company Omdia, doesn't pull any punches. "The UK is just tinkering at the edges of this, just like EVs and giga factories the UK is slow and ponderous," he says.

To be fair, things have been happening. Laura Foster, associate director for tech and innovation at techUK, praises initiatives such as ChipStart, a £1.3 million startup incubator from semiconductor industry accelerator Silicon Catalyst. She also applauds the £11 million each given to innovation and knowledge centers in Southampton and Bristol universities for photonics and compound semiconductor development. "Both of these initiatives will help the UK develop its capability in new and emerging modalities of semiconductors that will become an increasing part of the semiconductor market," she says.

We've also seen [9]the UK partner with Canada to build stronger supply chains for semiconductor technology.

In the June spending review this year, the government also carved out cash for a National Semiconductor Center as part of a new industrial strategy, which techUK had asked for in May 2024. This will be a key umbrella body, she says.

However, techUK also still thinks there's more to be done. "The consensus across industry is that, when compared to other nations, actioning of the UK's National Semiconductor Strategy has been unfortunately slow," says Foster. "Although the UK doesn't have the same market share as the United States, China, or Taiwan, we still have world leading capability across design, compound, and chips made for photonics and quantum. The UK will require bolder action and renewed priority, especially if we are to achieve its complementary ambitions in AI and quantum."

This is why techUK put forward its own plan for the UK chip sector, called the [10]UK Plan for Chips , in May 2024. Among other things, it called for more access to investment and more leadership.

Overall, the general consensus seems to be "lovely starter, now where's the main course?" Experts have broken down some key items on the collective to-do list.

More infrastructure

The national semiconductor strategy acknowledged that a lack of prototyping facilities has been a key weakness in the UK, and Foster still thinks this is an issue. "Pilot lines and prototyping environments are in short supply, which impedes scaling and investor confidence," she warns, calling for more onshoring of such facilities. "What if the UK became one of the best places in the world to prototype and pilot silicon photonics?" This is an area where government intervention is critical, she adds.

Grow the clusters

The UK's regional clusters (areas like South Wales where universities and private companies collaborate) are critical to industry growth. Making more of them is a common sentiment. Today they each concentrate on different capabilities, from compound semiconductors to advanced packaging and photonics. It's time to unify, say experts.

"The UK must take advantage of its clusters of semiconductor expertise across the country," says Raj Gawera, interim CEO at Compound Semiconductor Applications (CSA) Catapult. Innovate UK set this up in the South Wales cluster to work with universities and private industry to commercialize compound semiconductor tech.

Clusters are an economic driver and a source of jobs. We must do more with them, he says. "We must also do better at connecting clusters, to form what we call a 'super cluster', which will significantly increase innovation and growth," he says. "This will be part of the remit of the new UK Semiconductor Center."

Skill up

The skills gap is an area where the government must still do work. It faces tough competition, warns Illsey. "The vast majority of chips are made in Taiwan, so the expertise is in the Far East," he says. "The U.S. are trying to bring chip manufacturing back to the U.S. at scale, so the skills are in demand."

Gawera says that retention is also an issue. "Skills provision remains a significant barrier both nationally and globally," he says. While he's seeing positive signs of increased study in electronic engineering and related subjects, "keeping graduates in the UK in a globally competitive market and finding them good, well-paid jobs is a challenge." Perhaps that's where the clusters' job-nurturing capabilities can help. We could also do with a more welcoming immigration policy to source that talent from elsewhere, says Foster.

Foster supply chains

The UK Plan for Chips also re-emphasized the need for strategic supply chain partnerships with other countries. "We strongly welcome the delegations led primarily by InnovateUK and UKRI, but at times there has been a de-emphasis on this from within DSIT," Foster warns. "Helping companies identify and competitively source key partners as they move from design to piloting and manufacturing will help our most innovative businesses move effectively through the value chain."

She'd like to see the UK Semiconductor Centre extend beyond national borders, serving as a global interface between UK businesses and international companies.

Punch at its weight

One broad concern is that the UK might have an inferiority complex on the world stage.

"Outside of the US, the UK is the only provider of foundational, mainstream GPU technology worldwide," says Jake Kochnowicz, chief revenue officer at British IP and chip-design house Imagination Technologies, best known for the PowerVR GPU architecture and, more recently, IP for neural-network accelerators. "Despite punching well above its weight in chip design, the UK has yet to translate this strength into a policy framework that fully capitalizes on its competitive advantage."

Imagination Technologies initially praised the Semiconductor Strategy, he says. "However, in the face of substantial, coordinated support being provided by other regions to their domestic industries such as the EU's Chips Act, the US CHIPS and Science Act, and Japan's extensive semiconductor subsidies, the UK's approach risks falling short of the competitive benchmark."

Show me the money

Kochnowicz points to the change in government since the strategy was introduced (and subsequent alteration of fiscal policy) as an issue that has stopped the government driving the investment it needs.

[11]Softbank bets $2 billion on Intel having a future

[12]TechUK demands that Britain's chip strategy is crisped up

[13]Qualcomm set to move in on UK-listed chip IP biz Alphawave

[14]Arm juices mobile GPUs with neural tech for better graphics

"Looking ahead, the government needs to consider how it works with established companies, like Imagination, who are already harnessing the UK's engineering capabilities and strengths in IP and design. For companies like Imagination, clear commitments on R&D funding, targeted tax incentives, and support for international collaboration and talent access are essential."

Just over two years on, the verdict from at least some industry players seems clear: the UK government started well, and has made a laudable effort to put some foundational work in place for the UK processor sector. But in such as competitive market, there's no time to take your foot off the pedal. ®

Get our [15]Tech Resources



[1] https://www.theregister.com/2025/08/19/uk_tech_body_semiconductor_report/

[2] https://siliconangle.com/2025/08/13/report-u-s-government-placing-trackers-ai-server-shipments-enforce-chip-export-controls/

[3] https://www.theregister.com/2025/08/19/uk_tech_body_semiconductor_report/

[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/personaltech&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2aKXxGdJAbqbT_UXxyh7N7AAAAII&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0

[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/personaltech&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aKXxGdJAbqbT_UXxyh7N7AAAAII&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0

[6] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/personaltech&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aKXxGdJAbqbT_UXxyh7N7AAAAII&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0

[7] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/personaltech&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aKXxGdJAbqbT_UXxyh7N7AAAAII&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0

[8] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_onprem/personaltech&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aKXxGdJAbqbT_UXxyh7N7AAAAII&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0

[9] https://www.canada.ca/en/national-research-council/news/2025/07/canada-and-uk-partner-to-build-a-stronger-semiconductor-supply-chain.html

[10] https://www.techuk.org/resource/tech-uk-s-report-uk-plan-for-chips-a-new-blueprint-is-now-live.html

[11] https://www.theregister.com/2025/08/19/softbank_intel_investment/

[12] https://www.theregister.com/2025/02/17/techuk_semiconductor_strategy/

[13] https://www.theregister.com/2025/04/02/qualcomm_alphawave/

[14] https://www.theregister.com/2025/08/12/arm_bringing_neural_acceleration_to/

[15] https://whitepapers.theregister.com/



Adorable

Anonymous Coward

It's so cute when the UK thinks it can sit at the grown-ups' table!

45RPM

Nascent? Hardly. It was established, and potentially world beating, with companies like Inmos and ARM (also Ferranti, Plessey, Marconi and more besides). What did we do with them? We failed to invest and then threw them away.

It’s a bit like racists complaining about immigrants coming over here and diluting our culture. It’s the established Brits who are ruining our own culture and our own businesses. We don’t care enough to look after them. We don’t deserve to have nice things. We grumble. We moan. We coulda woulda shoulda. And we do nothing except self destruct.

Headley_Grange

To be fair, a lot of the capability you reference didn't disappear primarily due to lack of investment. Ferranti went bust because of its catastrophic acquisition of ISC. The semis were bought by Plessey but eventually ended up with US owners. Plessey was sort of bought out by GEC/Siemens then a few years later GEC/Marconi went bust after spending Arnie Weinstock's cash mountain on massively over-valued internet companies during the dot-com bubble.

The article describes the problem pretty well. It's a long-term strategic investment that requires sympathetic support coupled with firm guidance from government and supportin policies in planning, taxation, education, transport, housing, strategic protection to name a few. On the other hand, the UK stock market demands either quarter-on-quarter profit growth or magic beans (today it's AI) if your share price isn't to buckle and leave you open to takeover. New UK semiconductor manafacturing wouldn't make profit for a long time.

Vestas

...and just to finish off your history Racal spun off (demerged) Vodafone & Chubb, 26% of Camelot, then sold the rest of the defence businesses at vastly inflated prices to foreign (French) owners. Racal MESL was an early casualty but the sale of Racal Telecom to Global Crossing for a billion at the peak of the dot-com bubble was incredibly good business as it'd have gone the same way as Marconi.

£1000 invested in Racal in 1961 (when Harrison got them listed on the stock market) would have been worth £14.5 million in 2000 when he retired. Nice returns & what happened to the remains of the defence businesses was inevitable given UK govt spending in the late 90s.

45RPM

It's true. But a lot of the time, it happened because the UK government didn't protect the capital of this country. We allow our businesses to be bought out by foreign investment, we allow the profit to be funnelled abroad. Other countries don't - they protect their businesses, they see them as being in the national interest. The only time UK government does anything is if it looks like there might be mass unemployment in an area, and hence loss of votes, and sometimes not even then.

Why isn't ARM a British owned business anymore? It wasn't in any trouble - it was just that SoftBank saw the potential and we didn't. Why was Inmos sold to STM? It wasn't in any particular trouble - but STM wanted to strengthen their semiconductor technology portfolio. And it isn't just chip manufacture. Rowntree was fundamentally sound. ICL, whilst in financial difficulty, could have been seen as strategic and protected - but it wasn't. Similarly, it was an act of vandalism to sell Reuters to Thomson. or Cadbury to Kraft. We don't deserve nice things. We throw them away.

But you're absolutely right - so have a thumbs up Headley_Grange and Vestas. Nice summary.

elsergiovolador

The real problem isn’t just that Britain sells its best companies - it’s that the culture is designed to. The system makes sure working people never progress: twenty years in a firm and you’re exactly where you started, while the harder you work the bigger the shareholders’ yachts get and you might get an mp3 player and chocolate. Why build ARM into a national champion when you can flog it abroad and keep everyone in their lane? In the UK, mediocrity isn’t an accident - it’s the point.

Wishful thinking

may_i

There's a mistake in the start of the article. The phrase "with tiny die sizes" should be "with tiny feature sizes" the actual size of the die is dependent on both the complexity of the circuit and the size of the transistors and other die components. Some components simply can't be made smaller due to physics. Smaller transistors however translate to higher speeds and lower power requirements. The size of a die is largely irrelevant.

The "more infrastructure" part of the wish list is unrealistic. If you want to be able to prototype new designs which rely on very fast and very small transistors, you need the lithography capabilities for it. If you're going to invest in EUV lithography capabilities, there's no way you'll recoup the investment just doing prototypes.

Skilling up by trying to get talent from abroad to move to the UK is a non-starter. As the article states, there is already a shortage of skills worldwide. Even if you could attract scientists and engineers to move to the UK, they're not likely to stay and it will take very attractive salaries to get people to move in the first place. The most likely outcome is that foreign talent won't stay and will simply return home once they've improved their skills at the expense of the company which hired them.

In the end, playing catch up with the established players who have both manufacturing and design capabilities is not something that can be done in any short timescales. Fabless chip designers don't create that much in the way of jobs either. The UK missed the boat three decades ago. No amount of wishing and optimistic plans are going to change that.

Mindset

elsergiovolador

We could also do with a more welcoming immigration policy to source that talent from elsewhere, says Foster.

Welcoming immigration policy’ is code for: we’d rather import engineers from poorer countries, pay them barely enough to survive, and stack them twelve to a kitchen, than raise wages for locals. It’s the same feudal logic as always - a permanent underclass of grateful, disposable talent to prop up managers who spend more on their landlord portfolios than on payroll. Britain doesn’t have a skills shortage, it has a wages shortage. Dressing that up as ‘openness’ is just cheap-labour colonialism with a rainbow lanyard.

"you have to keep it up"

Pascal Monett

That's what she said . . .

The bureaucracy is expanding to meet the needs of an expanding bureaucracy.