Oxford-based chip startup Salience Labs has raised $125m (£94m) from investors, including AMD and Intel, as the race to build AI data centres puts growing pressure on electricity supplies.
The funding round, led by Bullhound Capital, is set to help the British firm scale production of its photonic chips, which use light rather than electrical signals to move data between computing systems.
The technology is designed to tackle the vast amounts of energy required to run increasingly powerful data centres, one of the biggest bottlenecks facing the AI industry.
AMD Ventures, Intel Capital, SBI Group, Hudson River Trading and the British Business Bank also participated in the Series B round.
Salience plans to use the money to expand its manufacturing capacity and strengthen its presence in Silicon Valley, as it looks to supply the infrastructure underpinning the global AI boom.
Unlike conventional networking equipment, which relies heavily on electrical signals, the company’s optical circuit switching (OCS) technology keeps data travelling as light with the aim of reducing power consumption and delays. The approach could allow data centres to devote more of their limited electricity supply to running AI models rather than moving information between chips and servers.
“OCS will play a critical role in AI datacenter networking”, said Vaysh Kewada, chief executive and co-founder of Salience Labs. “As customers roll out optical links in their datacenters, high-performance optical switches are needed to deliver the token economics that workloads demand.”
UK’s AI ambitions face an electricity bottleneck
The investment comes as Britain’s ambitions to become a global AI hub are hindered by high electricity costs and bottlenecks in accessing the national grid.
Demand for data centre connections has surged as technology companies race to secure enough power for AI infrastructure.
According to figures cited in a recent Ofgem consultation, the UK’s electricity connection queue tripled from 41 gigawatts to 125 gigawatts between November 2024 and June 2025.
The regulator has proposed tougher requirements for data centre developers to access the queue, including substantial upfront financial commitments, to prevent speculative projects from blocking access to the grid.
Meanwhile, some developers are considering building their own power generation facilities to avoid lengthy connection delays. The pressures have made technologies capable of reducing energy consumption increasingly attractive to investors and the world’s largest chipmakers.
Salience’s latest funding also highlights growing interest in British semiconductor companies developing the infrastructure needed to support AI, rather than the models themselves.
Earlier this year, Oxford-founded chip startup Fractile secured $220m from investors, including Peter Thiel’s Founders Fund, as it develops technology designed to make AI computing faster and cheaper.
The backing from AMD and Intel provides Salience with support from two of the industry’s biggest semiconductor players as it moves towards commercial-scale manufacturing.
Callum Stewart, principal at Bullhound Capital, said the startup had developed technology capable of meeting both the immediate and longer-term needs of data centre operators.