Anthropic, the San Francisco-based company behind the Claude AI model, is reportedly in discussions with Fractile, a London-based semiconductor startup, to secure a supply of custom chips for AI inference. The potential partnership is part of Anthropic's broader strategy to reduce its heavy reliance on Nvidia's hardware and gain more control over its technical infrastructure.
According to industry reports, the talks are still at an early stage and no binding agreement has been signed. However, the interest from a major AI player underscores the growing importance of specialised chip design in the race to deliver faster and cheaper AI compute power. Anthropic, which has received multi-billion-dollar investments from Amazon and Google, currently depends on Nvidia's H100 units alongside custom processors from its cloud partners.
A Shift in Semiconductor Strategy
The high market price and limited availability of Nvidia's industry-standard chips have squeezed profit margins for AI developers, prompting firms to explore alternatives. This trend is not limited to Anthropic; tech giants like Microsoft and Meta are also moving away from general-purpose chips in favour of internal or boutique designs. A deal with Fractile would make it Anthropic's fourth major chip supplier, joining Nvidia, Google, and Amazon.
Fractile, founded in 2022 by Oxford PhD Walter Goodwin, has attracted attention for its unconventional approach to processor design. Unlike standard chips that constantly shuttle data between the processor and separate memory modules, Fractile's "memory-compute fusion" architecture keeps data directly on the chip using static random-access memory (SRAM), which does not need to be refreshed. The startup claims this method can run large language models up to a hundred times faster than existing hardware while lowering operational costs by 90%.
While these performance claims are impressive, the technology is still in development. Fractile has not yet launched a commercial product, and its specialised chips are not expected to be ready for full-scale data centre deployment until 2027. Despite the long timeline, the startup is reportedly in negotiations to raise $200 million (€170.5 million) in funding at a valuation exceeding $1 billion (€853 million).
The potential partnership highlights the growing significance of the UK's semiconductor sector on the world stage. As the UK fits into Europe's integrated supply chain network, this deal could strengthen the continent's position in the global AI hardware market. The discussions also come at a time when European markets are closely watching geopolitical tensions that could affect supply chains, such as the impact of stalled US-Iran talks on oil prices.
According to market analysts, the ability to deliver faster and cheaper compute power is becoming the defining factor in the AI race. If Anthropic secures a deal with Fractile, it could gain a competitive edge by reducing operational costs and improving model efficiency. However, the long timeline for Fractile's commercial product means that any benefits are still years away.
For now, the talks remain exploratory, but they signal a clear shift in the AI industry's approach to hardware. As companies like Anthropic seek to diversify their supply chains, European startups like Fractile could play an increasingly important role in shaping the future of AI computing.
