Recommended Articles
Fundraising

Salience Labs Raises $125M Series B to Scale Photonic AI Data Center Switching

By Angeled · November 2, 2026 · Fundraising
Salience Labs Raises $125M Series B to Scale Photonic AI Data Center Switching
Salience Labs Raises $125M Series B to Scale Photonic AI Data Center Switching

Salience Labs, a photonics company spun out of the University of Oxford and the University of Münster, has closed an oversubscribed $125 million Series B funding round to scale its all-optical switching technology for artificial intelligence data centers. The round was led by Bullhound Capital, whose principal Callum Stewart joins the company's board of directors.

New investors in the round include AMD Ventures, Intel Capital, SBI Group, Hudson River Trading, and the British Business Bank. Existing shareholders Applied Ventures, Oxford Science Enterprises, Cambridge Innovation Capital, and Silicon Catalyst Ventures also participated. The company says the capital will accelerate deployment of its optical circuit switching (OCS) technology, scale volume manufacturing, and expand its presence in Silicon Valley.

The bottleneck inside the AI data center

Modern AI training runs do not happen on a single computer. They happen on clusters of thousands of graphics processing units (GPUs) spread across many racks, all of which must behave like one machine. The connections between those GPUs — the network — often determine how fast a model trains and how much power the whole system consumes.

In today's data centers, data typically travels between racks as light through optical fiber, but it is converted into electrical signals every time it passes through a switch, then converted back into light again. Each conversion costs time and energy. As clusters grow larger and models grow more demanding, that repeated conversion becomes a bottleneck: latency rises, power bills climb, and the network struggles to keep pace with the compute it serves.

Optical circuit switching takes a different approach. Instead of converting light to electricity and back, an OCS switch routes the light itself, keeping data in the optical domain as it moves between connections. Salience Labs describes its architecture as a fully integrated, all-optical switch designed to connect thousands of GPUs across multiple racks, targeting the "scale-up" networks that bind accelerators into tightly coordinated systems.

What Salience Labs has built

Salience Labs was founded in 2021 by Vaysh Kewada and photonics physicist Johannes Feldmann, building on more than a decade of silicon photonics research at Oxford and Münster. Silicon photonics is the practice of building optical functions — guiding, switching, and amplifying light — directly onto semiconductor chips, so they can be manufactured with chip-industry processes rather than assembled by hand.

The company announced a 32-port all-optical switch in March 2026 and then demonstrated a 64-port all-optical photonic switch live at ECOC 2026, the European Conference on Optical Communication. The company describes that demonstration as an industry first for silicon photonics OCS. The Series B will fund work toward 128-port and 256-port switches, which would extend the technology to larger and more demanding deployments. Salience is also scaling manufacturing with partners including Keysight Technologies and Tower Semiconductor.

The strategic interest of the investor list is hard to miss. AMD Ventures and Intel Capital are the venture arms of two of the largest chipmakers in the world, both of which sell into the same AI infrastructure buildout. Their participation, alongside a UK state-backed investor in the British Business Bank, signals that photonic networking is moving from research curiosity to a contested layer of the AI stack.

Why the market is paying attention

The commercial stakes are large. Dell'Oro Group, cited in Salience's announcement, forecasts that cumulative spending on data center switches deployed in AI back-end networks will approach $1 trillion over the five years from 2026 to 2030, with scale-up networking expected to account for more than half of that market by 2030. Even if only a fraction of that spending shifts toward optical switching, the prize for a company that can manufacture it at volume is enormous.

The timing also matters for founders and investors watching infrastructure. The first wave of AI investment went into the chips themselves — GPUs and accelerators. The current wave is going into everything around them: power, cooling, interconnects, and networking. Each of these layers has its own physics constraints, and each is attracting specialist startups. Salience's bet is that the switching layer, long treated as a commodity, becomes a performance differentiator as clusters scale.

For founders, the lesson is a familiar one in deep tech: the most valuable company in a stack is not always the one selling the headline component. It is often the one that removes the constraint everyone else runs into next. For investors, the round is a reminder that hardware investing has returned to fashion — but only for teams with a demonstrable product, manufacturing partners, and customers whose pain is measurable in latency and watts.

Angeled.com connects founders with investors.

Business Wire: "Salience Labs Secures $125 Million in Series B Funding to Scale Photonic Networking Chips for AI Datacenter Infrastructure"
Optica Optics & Photonics News: "Salience Labs Raises US$125 Million To Go All-Optical"
Pulse 2: "Salience Labs Raises $125 Million Series B Led By Bullhound Capital For AI Data Center Photonic Networking Chips"
Unite.AI: "Salience Labs Raises $125 Million Series B to Scale Optical Switching for AI Datacenters"

Read Next