A top executive at one of the UK‘s hottest AI startups has made the bold claim that it has the potential to be Britain's first one trillion dollar company. Carmen Alfonso Rico is a VC-turned-angel investor-turned operator. Rico, who started her own VC fund, Cocoa Ventures, in 2021, joined UK chip startup Fractile, a UK startup that designs chips for AI inference, the process of running an AI model, in June this year.
Cocoa Ventures, which has raised a $17m and $23m fund, invested in Fractile in a 2024 funding round. Speaking on the Tech.eu podcast, Rico, who has joined Fractile as VP of business operations, issued the bold claim about its potential to be a one trillion dollar company, as she set out Fractile’s potential as it looks to take on chip titan Nvidia.
Rico said that Fractile had "a unique chance to build Britain's first one trillion dollar company" should it win the AI hardware war. However, the startup, which has yet to launch a commercial product, still has some way to go to even become a multi billion dollar company, let alone a trillion dollar company.
Fractile, which is also backed by Accel and Peter Thiel's Founders Fund, is reportedly raising $600m at a $5.6bn valuation, according to Bloomberg, and has also reportedly signed a deal to sell its chips to Anthropic. Rico talks about why she decided to take the full-time role at Fractile, which she is undertaking at the same time as overseeing Cocoa Ventures, although a Cocoa Ventures executive will now run its day to day operations.
Rico pitched it as a move which will benefit Cocoa Ventures and its other portfolio companies, highlighting the experience and access the operator role will grant her.
She said: "I have a better network. I am going to support the portfolio because of the place that Fractile operates in the world.”
Rico said she had no pushback from portfolio companies to the move, saying she is always available to them. Founded in 2022 by Oxford PhD Walter Goodwin, Fractile’s architecture keeps data directly on the chip, unlike standard chips that must constantly shuttle data between the processor and separate memory modules.
It says this method can run LLMs up to a hundred times faster than existing hardware while lowering operational costs by 90 per cent. Fractile's specialised chips are expected to be ready for data centre deployment in 2027.
Would you like to write the first comment?
Login to post comments
Facts Only
* Carmen Alfonso Rico is a VC-turned-angel investor-turned operator.
* Rico joined UK chip startup Fractile as VP of business operations in June this year.
* Cocoa Ventures, founded in 2021, invested in Fractile during a 2024 funding round.
* Rico claimed Fractile has the potential to be Britain's first one trillion-dollar company.
* Fractile's architecture keeps data directly on the chip.
* This method reportedly allows LLMs to run up to a hundred times faster and lowers operational costs by ninety percent.
* Fractile’s specialized chips are expected for data center deployment in 2027.
* Fractile is backed by Accel and Peter Thiel's Founders Fund.
* Fractile is reportedly raising $600 million at a $5.6 billion valuation.
* There is a reported deal to sell Fractile's chips to Anthropic.
Executive Summary
Full Take
The narrative relies on positioning technological breakthrough as an inevitable route to unprecedented economic scale, framed by the specific context of the AI hardware war against Nvidia. The claim of becoming a one trillion-dollar company serves as a potent aspirational anchor, leveraging the massive cultural and financial gravity of current AI development. The mechanism presented—a novel chip architecture enabling tenfold speedups and ninety percent cost reduction for LLMs—presents a compelling technological leap that justifies the ambition, even if the timeline to a trillion dollars remains highly speculative, especially given Fractile has not yet launched a commercial product.
The focus on Rico's network and ability to benefit her portfolio suggests that access and operational experience are treated as primary drivers of future success, implying that the infrastructural advantage is less critical than the strategic positioning achieved through high-level investment and operational roles. This structure echoes patterns where expertise or access is presented as a shortcut around pure market validation or technological execution risk. The underlying assumption is that a foundational technological shift will automatically translate into dominant economic power, ignoring the significant friction involved in scaling hardware innovation from lab to trillion-dollar enterprise.
What are the missing variables regarding market adoption curves versus technological readiness? If the performance metrics (speed and cost reduction) hold, what empirical data exists on the sustained scalability of this advantage across diverse real-world deployment scenarios? Does the ambition mask a potential diversion from focusing on demonstrable milestones necessary for long-term, sustainable valuation rather than hyperbolic projections.