Salience Labs, an Oxford-based maker of photonic networking chips, has closed an oversubscribed $125 million Series B round to scale its all-optical switches for AI data centers. Why it matters: as GPU clusters grow, moving data between chips — not raw compute — is becoming a central cost and speed bottleneck, and Salience is betting light can beat conventional electronic switching.
October 14, 2026 — The round was led by Bullhound Capital, whose principal Callum Stewart joins the company’s board. New investors include AMD Ventures, Intel Capital, SBI Group, Hudson River Trading and the British Business Bank, with existing backers Applied Ventures, Oxford Science Enterprises, Cambridge Innovation Capital and Silicon Catalyst Ventures also taking part.
What the company builds
Salience develops optical circuit switching (OCS) technology: a fully integrated, all-optical switch architecture designed to link thousands of GPUs across multiple racks. The idea is to keep data in the optical domain rather than repeatedly converting signals back and forth between light and electricity, which the company argues adds latency and power draw in large AI clusters.
The company recently demonstrated a 64-port all-optical photonic switch at ECOC 2026, which it describes as a first for silicon photonics OCS, and says 128-port and 256-port versions are in development. Its technology builds on more than a decade of silicon photonics research; the company was founded in 2021 and spun out of work at the University of Oxford and the University of Münster.
Where the money goes
Salience says the new capital will speed up deployment of its switching technology, expand volume manufacturing, and grow its U.S. footprint in Silicon Valley, including a second West Coast location. It also plans to keep working with partners including Keysight Technologies and Tower Semiconductor as it moves toward high-volume production.
The release cites a Dell’Oro Group forecast that cumulative spending on data center switches for AI back-end networks could approach $1 trillion over the next five years, with scale-up networking accounting for more than half of that market by 2030.
Why it matters
For anyone following the AI build-out, this is another sign that the money is moving down the stack. The headline race is about bigger models and more GPUs, but those GPUs only earn their cost when data moves between them fast enough. Having both AMD Ventures and Intel Capital in the same round is a useful signal: two rival chip camps are backing the same networking supplier rather than treating interconnect as an afterthought. The test ahead is manufacturing — photonic demos are one thing, shipping switches at data center scale is another — and Salience is explicit that this round is largely about crossing that gap.
Frequently asked questions
How much did Salience Labs raise, and who led the round?
The company raised $125 million in an oversubscribed Series B led by Bullhound Capital, with participation from AMD Ventures, Intel Capital, SBI Group, Hudson River Trading, the British Business Bank and several existing investors.
What does Salience Labs actually make?
It makes photonic — light-based — networking chips and optical circuit switches designed to connect thousands of GPUs across racks in AI data centers with lower latency and power use than conventional electronic switching.
What will the funding be used for?
Deploying its optical switching technology, scaling volume manufacturing, developing higher-port-count switches, and expanding its presence in Silicon Valley and other markets.
Sources: Business Wire, SiliconANGLE, Tech.eu