The UK can invent world-class chips. Scaling them is the hard part

We spent the end of September at Microelectronics UK, the trade show at ExCeL London that gets the whole UK chip world into one building. The talk we haven’t stopped thinking about since was Addressing the Scale Up Challenge, from Andy McLean, chief executive of the UK Semiconductor Centre [1]. McLean has spent his career in semiconductors and seen most of the world’s chip ecosystems up close, and his verdict on the UK started as a compliment. “There is no better place in the world for research and innovation, everything from fundamental research through to applied science.”

But the compliment came with a problem attached. The UK produces world-class chip science, IP and design talent, then watches too much of it scale somewhere else, or scale here on someone else’s money. The UK Semiconductor Centre is the national body set up to change that, and McLean sums up its mission in one word: scaling. His talk answered one question: why has that been so hard?

The scale-up capital mostly isn’t British

Start with who actually funds UK companies right now. In the first half of 2025, 66% of the venture capital invested into UK startups came from international investors. US funds alone accounted for 43% (Tech Nation Report 2025) [2]. You see the same pattern in deal counts: in 2024, UK equity deals with at least one overseas investor outnumbered deals funded only by domestic investors, 625 to 574 (British Business Bank Small Business Equity Tracker 2025) [3].

McLean was careful not to paint foreign money as the enemy. “If we can scale startups with FDI, no problem. But wouldn’t it be great if more of it came from the UK investment community?”

And his fix runs both ways. One half is showing UK investors how they can prosper by backing UK semiconductors. The other half is helping startups understand what it takes to land that investment in the first place.

The deep tech money has started to show up

If you’d written off UK deep tech as unable to attract serious cheques, this year says otherwise. Fractile, a UK AI chip startup, raised a $220m Series B in May 2026 at a $1bn valuation [4]. Oxford Quantum Circuits closed a £260m Series C in June 2026, the largest private round ever raised by a European quantum company, and £100m of it came from the British Business Bank [5]. That’s UK institutional capital turning up at later stage rounds.

And it’s not like the UK is short of venture capital overall. In 2025, UK startups raised more venture capital than startups in France, Germany and Switzerland combined. That makes the UK comfortably Europe’s largest venture market (Dealroom/HSBC UK Innovation Update, Q3 2025) [6]. The pipeline deep tech grows from is strong too: university spinouts raised £1.86bn across 243 deals in 2024, which was 17% of all UK equity investment that year (British Business Bank) [3].

So the research is here, and the money is here. The gap, in McLean’s telling, sits between them. Not enough of that capital knows the sector well enough to back its scale-up rounds with conviction. That’s why so many of those rounds are led from abroad.

What government can do, and what it can’t

The state has moved this year. In June 2026 the government announced a £1.1bn AI Hardware Plan [7]. The part McLean singled out is the Advanced Market Commitment: up to £150m for the government to buy novel inference chips directly from UK companies, with a further procurement of up to £250m to follow [7]. For a hardware startup, a committed lead customer is often the difference between a fundable story and an unfundable one. Here, the government is putting its hand up for the job.

He was equally clear about the limits of public money. “Government can’t fund to the extent that’s necessary, but it can seed activity and signal to the private investment community that they should be taking notice of what’s going on in semiconductors.”

The same pragmatism applies to infrastructure. Asked when the UK would get its own advanced fab (semiconductor fabrication plants), McLean didn’t hedge. “When is the UK going to have an advanced fab? My answer is pretty straightforward: never. We don’t need one.” Instead, the UK Semiconductor Centre signed an agreement with Rapidus in June 2026 that gives UK companies access to the Japanese consortium’s 2-nanometre manufacturing, expected from late 2027 [8]. In other words: design the chips here, make the leading-edge silicon with allies, and spend the capital on what the UK is actually good at.

What founders raising in deep tech should take from this

First, the numbers involved. McLean’s own estimate from the stage, and it’s his estimate rather than a published statistic, is that getting a semiconductor product to prototype takes £100m to £200m. If you’re sizing a raise in this sector against software norms, that’s the corrective. Rounds that once topped out at £10m now run to hundreds of millions, so capital strategy, milestones and investor conversations all need planning at that scale.

Second, fit matters more here than almost anywhere. A generalist investor can’t easily evaluate technical risk in photonics or advanced packaging, so time spent finding investors who already understand the sector is rarely wasted.

And third, the connective tissue is still being built. “We have incredibly strong communities throughout the UK. They don’t talk enough,” McLean told the room. His centre’s answer is to get academia, government and industry into the same room. For an individual founder, the practical version is smaller: the investors who understand your sector are out there, in the UK as well as outside it, and the job is finding them early, while you still have time to choose. It’s the same gap ThatRound spends its days on, relevance between founders and the investors who actually understand what they’re building.

The UK has spent decades proving it can invent the future of computing. The next few years, if McLean is right, are about proving it can own more of the upside. The research base isn’t the question. Whether UK capital shows up for it is.

References

1. Addressing the Scale Up Challenge | Microelectronics UK 2026, 29 September 2026
2. Tech Nation Report 2025 | Tech Nation
3. Small Business Equity Tracker 2025 | British Business Bank
4. UK AI chip startup Fractile raises $220M | Tech.eu: https://tech.eu/2026/05/13/uk-ai-chip-startup-fractile-raises-220m-to-tackle-the-growing-inference-bottleneck/
5. OQC Raises £260 Million in Series C | The Quantum Insider: https://thequantuminsider.com/2026/06/03/oqc-raises-260-million-in-series-c/
6. UK Innovation Update Q3 2025 | Dealroom/HSBC
7. Government announces UK AI Hardware Plan | techUK: https://www.techuk.org/resource/government-announces-uk-ai-hardware-plan-techuk-reaction.html
8. Rapidus signs MoU with UKSC | Electronics Weekly: https://www.electronicsweekly.com/news/business/rapidus-signs-mou-with-uksc-2026-06/