Fractile raises $720 million Series C at $7 billion valuation for AI chips

UK-based AI chip startup Fractile raised $720 million in a Series C funding round led by Thrive Capital, Founders Fund, Lightspeed, and Redpoint, elevating its valuation to $7 billion. Fractile is developing custom chips with stacked memory designed to accelerate AI model inference by up to 10-fold and lower costs compared to market leader Nvidia. Backed by the UK Sovereign AI Fund, Fractile plans to launch its first commercial hardware products in 2027.

Series C funding details

• ▪London-based AI chip startup Fractile raised $720 million in a Series C funding round at a $7 billion valuation
• ▪Fractile's Series C funding round was led by venture capital firms Thrive Capital, Founders Fund, Lightspeed Venture Partners, and Redpoint
• ▪Fractile's $7 billion Series C valuation represents an increase from its $1.1 billion valuation during its $220 million funding round in May 2026
• ▪Investors participating in Fractile's $720 million Series C round included CapitalG, Conviction, BlackRock, DST Global, Elad Gil, Balderton, SV Angel, the British Business Bank, and the U.K. Sovereign AI Fund, supporting domestic U.K. artificial intelligence infrastructure

Fractile chip architecture and performance claims

• ▪Fractile is developing artificial intelligence chips with custom memory built directly into its chips to speed up inference tasks
• ▪Thrive Capital investor Philip Clark stated that Fractile's chips are particularly well suited for artificial intelligence agents operating independently for hours or days
• ▪Fractile is developing a full-stack architecture that includes a custom machine learning compiler alongside hardware to maximize model execution speed
• ▪Fractile founder Walter Goodwin claimed that building custom memory directly into Fractile's chips could increase artificial intelligence query response times 10-fold

Wider AI inference chip market context

• ▪Nvidia acquired chip startup Groq in 2025 and launched its own dedicated artificial intelligence inference chip in March 2026
• ▪Startups including Fractile, Axelera, Etched, MatX, and Olix are developing specialized semiconductors to process artificial intelligence inference faster and cheaper than Nvidia graphics processing units
• ▪Thrive Capital investor Philip Clark projected that the artificial intelligence inference market will eventually become at least as large as the market currently addressed by Nvidia

UK government AI strategy and funding goals

• ▪Britain made becoming an artificial intelligence superpower a central part of its economic strategy to retain jobs, investment, and value generated by technology
• ▪U.K. AI Minister Kanishka Narayan stated that the U.K. government is taking steps to ensure domestic scientific innovations generate technology, industry, and wealth within Britain