Fractile raises $720M to advance AI chip development
UK-based AI chip startup Fractile has secured $720 million in Series C funding, reaching a $7 billion valuation despite not yet shipping any production chips.
The London and Bristol-based startup closed the Series C round on November 11, 2026, approximately six months after its previous $220 million Series B round in May 2026, which valued the company at $1.1 billion. This latest investment represents more than a sixfold increase in valuation in roughly half a year.
Investor Confidence
The round was led by prominent venture firms including Thrive Capital, Founders Fund, Lightspeed Venture Partners and Redpoint. Other notable participants included CapitalG, the independent venture arm of Alphabet, as well as BlackRock and DST Global. The substantial investment underscores the market's enthusiasm for AI hardware alternatives despite Fractile's lack of commercial products.
Technology Focus
Founded in 2022 by Oxford researcher Walter Goodwin, Fractile specializes in AI inference chips. The company's innovative approach involves stacking DRAM directly onto the processor to reduce bandwidth limitations that slow down inference tasks. Fractile plans to manufacture its first commercial chips in 2027, targeting data center deployments that same year.
has a greater stake in the hardware that will underpin its future economy
Market Context
The funding reflects growing demand for alternatives to Nvidia's dominant position in the inference chip market. While Fractile has not yet finalized any customer agreements, there are indications of potential interest from Anthropic. The UK government has been actively supporting domestic AI manufacturing capabilities, positioning Fractile as part of this broader national strategy.
Fractile faces competition from other well-funded startups including Axelera, Etched and OLIX, all pursuing similar opportunities in the AI chip space. The company's success will depend on its ability to deliver on its 2027 manufacturing targets and secure customer adoption for its stacked-DRAM approach.