# UK government commits £1.1 billion to chip startups in bid to build domestic AI hardware sector > Source: > Published: 2026-08-21 12:01:06+00:00 Photo: Jim Osley / Wikimedia Commons / CC BY-SA 2.0 (https://creativecommons.org/licenses/by-sa/2.0) # UK government commits £1.1 billion to chip startups in bid to build domestic AI hardware sector The plan includes a national AI supercomputer, a £150 million advance purchase of inference chips from homegrown firms, and a target of capturing 5% of the global AI chips market by the early 2030s. The UK government just placed a very large bet that it can grow its own silicon. Announced during London Tech Week on June 8, the £1.1 billion ($1.47 billion) AI Hardware Plan is designed to do something Britain has struggled with for decades: turn promising chip startups into globally competitive semiconductor companies before they get acquired by foreign buyers or simply relocate. The centerpiece is a £750 million allocation for a national AI supercomputer, paired with £400 million dedicated to developing next-generation AI chips. But the most immediately interesting piece is a £150 million advance purchase commitment, where the government will buy inference chips directly from emerging UK firms. ## Where the money is going Two startups sit at the center of the initiative: Olix and Fractile. Olix, founded in 2024 by James Dacombe, builds photonic AI inference chips, a technology that uses light instead of electrical signals to process data. Together, Olix and Fractile had raised over £320 million ($440 million) before the government’s announcement. The government’s Sovereign AI Fund, launched in April 2026, has already put its money where its policy documents are. The fund participated in Olix’s $312 million Series B round, which valued the company at $3.3 billion. There’s also a new venture fund in the mix, led by Playground Global and backed by the British Business Bank with up to £150 million. Its mandate is to validate UK chip technologies. ## The strategic calculus The UK’s stated ambition is to capture roughly 5% of the projected $1 trillion global AI chips market by the early 2030s. If achieved, that would translate into approximately $50 billion in annual revenue. The advance purchase model tackles the classic chicken-and-egg problem facing hardware startups. Chip companies need large customers to justify manufacturing runs, but large customers won’t commit until the chips are proven at scale. By acting as a guaranteed buyer, the government absorbs some of that early risk. Britain has watched this movie before, with less happy endings. ARM, the country’s most famous chip company, was sold to Japan’s SoftBank in 2016 for $32 billion and later taken public in the US. DeepMind, the AI research lab, was acquired by Google in 2014. ## AI growth zones and talent pipelines Beyond direct funding, the plan introduces AI growth zones, designated areas where regulatory and planning processes are streamlined to accelerate the construction of data centers and semiconductor facilities. The timing of the announcement is deliberate. Global demand for AI inference chips—the processors that run trained models rather than training them—is growing rapidly as companies deploy AI applications at scale. By focusing on inference chips specifically, UK startups like Olix are targeting the segment of the market expected to see the fastest growth over the next decade. **Disclosure:** This article was edited by Editorial Team. For more information on how we create and review content, see our [Editorial Policy](https://cryptobriefing.com/editorial-policy/).