Salience Labs secures $125M to scale photonic chips for AI data centres

Salience Labs just pulled in $125 million in an oversubscribed Series B, and the money is going straight at one of AI’s messier plumbing problems: how to move data between GPUs fast enough, and without turning the network into a power hog.

Bullhound Capital led the round. AMD Ventures, Intel Capital, SBI Group, Hudson River Trading and British Business Bank joined in, alongside existing backers Applied Ventures, Oxford Science Enterprises, Cambridge Innovation Capital and Silicon Catalyst Ventures. Callum Stewart, a principal at Bullhound Capital, is also joining the board.

The company builds optical circuit switching technology for AI data centre connectivity. Salience says its all-optical architecture can link thousands of GPUs across multiple racks, and that matters because the compute side of AI keeps scaling while the network side gets pushed harder on performance, efficiency and power consumption. The pitch is straightforward: move traffic with light, not with the usual electronic bottlenecks.

This is not a clean-sheet science project anymore. Salience says the architecture is built on more than a decade of silicon photonics research and is aimed at high-volume manufacturing. The company recently demonstrated a 64-port all-optical photonic switch at ECOC 2026, which it calls an industry first for silicon photonics-based optical circuit switching. It is already developing 128-port and 256-port switches, while working with Keysight Technologies and Tower Semiconductor to scale manufacturing.

The funding will support deployment, product development and expansion. Salience also plans a bigger footprint in Silicon Valley, with a second West Coast location, plus entry into additional markets and growth in its engineering team. That is a familiar AI hardware story: the idea is elegant, the factory is the hard part.