A wave of funding and a new government hardware plan have put Britain's semiconductor industry back in the spotlight this year. We think the more interesting story lies beneath the headline numbers, which is where the UK's actual strengths lie within a global chip supply chain it doesn't quite control.
In early August, a two-year-old London startup called OLIX raised €270.5m in a Series B round that valued it at $3.3bn. The investor list included Arm - the Cambridge-founded chip designer whose architecture sits inside nearly every smartphone on the planet - alongside Hudson River Trading and Netflix co-founder Reed Hastings.
OLIX designs custom silicon chips for AI inference, combining SRAM architecture with photonic interconnects, and the round followed a $220m Series A only six months earlier.
We think this single deal captures the UK's semiconductor industry in 2026 more clearly than any policy document could, a country building on design strength rather than chasing chip factories it was never going to build.
The state of the sector
That funding round sits inside a semiconductor industry that has grown quickly by its own recent standards. A government-commissioned study published in June identified 705 UK semiconductor companies with 295 "dedicated" chip businesses and 408 with semiconductors as part of a wider offering, up from 623 at the previous baseline in 2024.
Dedicated companies generated an estimated £10.6bn in revenue and directly employed around 16,350 people in 2025, both up roughly 9% on the year before. Grants and fundraising across the sector rose 16% to £1.73bn, with newly identified companies securing a further £400m - weighted heavily toward design-led, early-stage firms in the East of England, London and the South East. Activity is concentrated across twelve recognised regional clusters, splitting fairly cleanly between design strength (Cambridge, London, Bristol, Southampton) and manufacturing depth (South Wales, Scotland, North East) - a split that mirrors government strategy almost exactly.
Optimism is real but not really universal.
Eighty-three per cent of firms surveyed expect growth over the next three years, and nearly half expect growth above 20% annually. Talent shortages, scale-up capital and energy costs were all flagged as ongoing constraints. Seventy per cent of dedicated companies remain UK-headquartered, and the vast majority are still SMEs rather than the household names most people associate with chips.
Design, not fabrication
The UK's National Semiconductor Strategy, published in 2023, was blunt about where the country's advantage lies, so rather than competing with Taiwan or South Korea on advanced silicon chip manufacturing, the £1bn strategy committed government funding to design, R&D and compound semiconductors - the three areas it identified as genuine UK strengths.
That doesn't mean the UK has no fabrication capacity at all.
Newport Wafer Fab, now owned by US firm Vishay Intertechnology after a forced sale away from its previous Chinese-linked owner, still makes wafers in South Wales, and Vishay committed a further £250m to expand the site in 2025. What the UK doesn't have is a leading-edge fab capable of the advanced nodes that power the newest AI accelerators, and the National Semiconductor Strategy explicitly chose not to chase one.
We still think that's a defensible bet, even if it wasn't universally popular at the time. Critics noted the £1bn figure looked modest next to the US's $52bn CHIPS Act and the EU's €43bn Chips Act, but money spent on design travels further than money spent chasing a fab the UK was unlikely to win outright.
Where the compound semiconductor cluster fits
South Wales is the clearest example of the design-and-materials strategy working in practice. The CSconnected cluster around Cardiff and Newport is the world's first dedicated compound semiconductor cluster, bringing together Cardiff University, IQE, Vishay and a growing base of startups around gallium nitride and silicon carbide materials - used in everything from electric vehicle power systems to 5G infrastructure.
Cambridge tells a related but different story: less about materials, more about the sheer density of chip design talent radiating out from Arm's Fulbourn Road headquarters. Together, the two clusters roughly map onto the UK's twin strengths, with design in the east and materials manufacturing in the west.
ARM's shadow, and the RISC-V question
ARM remains the UK's semiconductor industry success story by a wide margin. Its architecture underpins the vast majority of smartphone chips manufactured worldwide, and its willingness to invest directly in startups like OLIX shows the company acting as an active participant in the next generation of British chip design, not simply a licensing business collecting royalties from a distance.
ARM's dominance isn't unchallenged, though. RISC-V, the open-source instruction set architecture, has been steadily gaining ground as an alternative that doesn't require licensing fees at all.
Imagination Technologies, the Kings Langley-based chip designer better known for its graphics processors, has offered RISC-V CPU cores since 2022, positioning itself for a future where more of the UK's design work doesn't run through Arm's architecture by default.
We'd argue this is worth watching closely, because if RISC-V adoption accelerates, the UK's semiconductor industry stops being quite so synonymous with one Cambridge company, for better or worse, depending on how exposed British design firms are to Arm licensing revenue versus RISC-V's more distributed model.
None of this happens in a policy vacuum, the government's £1.1bn AI Hardware Plan, announced at London Tech Week in June, layered fresh money on top of the 2023 National Semiconductor Strategy, including a £750m national AI supercomputer, a £150m advance commitment to buy chips from British startups, and a new investment fund backed by the British Business Bank.
ChipStart UK, the strategy's earliest startup-facing program, has now run four cohorts and helped its alumni raise close to £100m in private investment.
Whether that's enough remains an open question.
Design strength and targeted policy have kept the UK relevant in a global semiconductor industry it doesn't otherwise control, but as AI compute demand reshapes the entire chip value chain, we're not convinced design alone is sufficient without a clearer route from promising startup to global scale.
That's the gap ChipStart and the AI Hardware Plan are explicitly trying to close but whether they manage it before the next funding cycle in Washington or Brussels is the question worth watching from here.
Also read: Advance Subscription Agreement (ASA): What UK Founders Need to Know
Editorial Note- This piece draws entirely on public data, government publications and reported funding announcements. We did not speak directly with any of the companies, investors or officials named, so please get in touch if that's you and you'd like to add anything.
Sources: DSIT, Semiconductor Sector Study 2026 (June 2026); DSIT, National Semiconductor Strategy (May 2023); GOV.UK, "A decisive shift to power British AI" (June 2026); EU-Startups (August 2026); DataCenterDynamics (August 2026); Innovation News Network (August 2026); The Conversation (July 2026); CEPA (June 2026); SiliconCatalyst.UK (March 2026); CSconnected; The Register (2023).
