British AI chip startup Fractile has raised $220 million (£165 million) in a major Series B funding round as global competition intensifies to build the next generation of AI infrastructure.
The London-founded company is developing specialized chips designed to dramatically improve AI inference — the process through which artificial intelligence models generate responses.
The round was led by Accel, Factorial Funds, and Founders Fund, the venture capital firm co-founded by billionaire investor Peter Thiel.
Additional investors included Conviction, Felicis, and 8VC.
The investment highlights growing interest in AI hardware startups focused on solving one of the industry’s biggest emerging problems: inference speed, cost, and scalability.
Why AI Inference Is Becoming the Next Big Battleground
Much of the AI boom has so far centered around training increasingly larger models such as ChatGPT. However, investors and researchers are now shifting attention toward inference — the stage where trained AI models actually generate outputs for users.
As AI systems become more advanced, they are being tasked with increasingly complex workloads that require longer reasoning chains and larger outputs.
According to Fractile, current hardware architectures are struggling to keep pace.
Founded in 2022 by Oxford-trained engineer Walter Goodwin, the startup believes the future limitations of AI will not come from model intelligence alone, but from how long it takes models to produce useful responses at scale.
“We bet everything on the logical conclusion: that the only way to truly unlock this latent value, to make speed viable at scale, was to radically re-invent the hardware that we run our frontier AI models on.”
Fractile Wants to Reinvent AI Chips From the Ground Up
Fractile is building custom AI chips designed specifically for high-speed inference workloads.
The company argues that today’s chips — particularly traditional GPU-based systems — face major bottlenecks involving memory bandwidth, energy consumption, and processing efficiency.
According to Goodwin, modern large language models can already generate outputs stretching to tens of millions of tokens while attempting to solve difficult problems.
However, many existing AI chips process outputs at roughly 40 tokens per second.
“At the roughly 40 tokens per second at which these models tend to run on existing chips, a single output of this length takes a month to complete.”
Fractile says its architecture redesigns how memory and power interact within AI systems, potentially enabling significantly faster inference while reducing energy costs.
This could become increasingly important as AI companies search for alternatives to expensive and power-hungry GPU infrastructure.
UK Government Calls Funding a “Vote of Confidence” in British AI
The UK government has pointed to Fractile’s funding round as evidence that Britain can produce globally competitive AI infrastructure companies.
UK AI minister Kanishka Narayan described the investment as:
“A strong vote of confidence in British AI.”
Narayan added:
“It shows that UK companies at the cutting edge are pulling in global investment while anchoring high value jobs and expertise here at home.”
He also emphasized that next-generation AI chips will play a major role in the country’s future economic growth and national security.
The comments come as the UK government prepares to unveil its broader AI hardware strategy later this year in an effort to strengthen Britain’s position in the global AI infrastructure race.
Rising Demand for Alternatives to Nvidia
Fractile’s raise comes amid growing pressure on dominant AI chip providers such as Nvidia.
While Nvidia’s GPUs remain central to today’s AI boom, concerns around supply constraints, rising costs, and energy usage have created opportunities for startups building alternative architectures.
Investors are increasingly backing companies focused on inference acceleration, which many believe could become the next major phase of AI infrastructure development.
The funding round also follows another major UK AI investment announcement: Isomorphic Labs recently secured $2.1 billion in Series B funding to expand its AI-powered drug discovery platform.
Together, these deals suggest growing momentum within the UK’s AI ecosystem as the country seeks to establish itself as a serious global player in artificial intelligence infrastructure.
Fractile Hiring Across Multiple Global Hubs
Following the funding round, Fractile is continuing to expand internationally.
The company is currently hiring across:
- London
- Bristol
- San Francisco
- Taipei
The new capital is expected to accelerate development of Fractile’s inference-focused AI hardware as competition in the sector rapidly heats up.
Final Thoughts
Fractile’s $220 million Series B funding round signals a major shift in AI infrastructure investment toward inference optimization.
As AI models grow more sophisticated and computationally demanding, the ability to generate outputs quickly, efficiently, and affordably is becoming one of the industry’s most critical challenges.
With strong backing from global investors and support from the UK government, Fractile is positioning itself at the center of the next phase of AI hardware innovation.
