Salience Labs, a company specialising in photonic solutions for AI data centre connectivity, has secured $30 million in Series A funding. This investment aims to advance the development of their optical switches, designed to enhance large-scale AI infrastructure.
The funding round was led by Cambridge Innovation Capital and Oxford Science Enterprises, with participation from Oxford Investment Consultants, former Dialog Semiconductor CEO Jalal Bagherli, Silicon Catalyst, the Goh Family Office in Singapore, and Arm-backed Deeptech Labs.
Salience Labs, founded in 2021 as a spin-out from the University of Oxford and the University of Münster, is developing an ultra high-speed multi-chip processor that integrates a photonics chip with standard electronics. This technology is scalable, capable of stacking up to 64 vectors into a beam of light, delivering massively parallel processing performance within a given power envelope.
The company’s proprietary amplitude-based approach to photonics results in dense computing chips operating at tens of gigahertz. This, combined with massively parallel performance, aims to enable exascale computing in a wide array of AI processes and applications.
Salience Labs employs a multi-chip design, with the photonic processing mapping directly on top of the Static Random Access Memory . This novel ‘on-memory compute’ architecture is inherently faster and can be adapted to the application-specific requirements of different market verticals, making it suitable for AI use-cases in communications, robotics, vision systems, healthcare, and other data workloads.
The technology has been designed from first principles for volume manufacture and is currently being fabricated with production-level foundries using standard CMOS processes.
Ian Lane, Partner at Cambridge Innovation Capital, stated that Salience Labs brings together deep domain expertise in photonics, electronics, and CMOS manufacture. He noted that their unique approach to photonics delivers an exceedingly dense computing chip without having to scale the photonics chip to large sizes.
Alexis Zervoglos, Partner at Oxford Science Enterprises, commented that Salience Labs offers a pioneering yet pragmatic solution to the widening gap between AI processing demand and semiconductor industry supply. By leveraging photonics in a novel way, the company is set to make unprecedented advances in computing, transforming existing AI applications and opening up new addressable horizons.
Jalal Bagherli, investor and former CEO of Dialog Semiconductor, remarked that the team at Salience Labs is world-class, combining commercial and technical acumen with a disruptive market vision. He highlighted their unique ‘on-memory compute’ architecture as ground-breaking, with the potential for breakthrough performance and power capability beyond what the established CMOS roadmap offers.
Vaysh Kewada, CEO and co-founder of Salience Labs, emphasized the increasing demand for faster chips to grow AI capability, noting that the semiconductor industry cannot keep pace with this demand. She explained that their proprietary ‘on-memory compute’ architecture combines the ultra-fast speed of photonics, the flexibility of electronics, and the manufacturability of CMOS, ushering in a new era of processing where supercompute AI becomes ubiquitous.
Salience Labs is headquartered in Oxford, UK, and funded by a portfolio of investors including Cambridge Innovation Capital, Oxford Science Enterprises, Oxford Investment Consultants, Silicon Catalyst, Deeptech Labs, and leaders from the global semiconductor industry, including former Dialog Semiconductor CEO Jalal Bagherli and the Goh Family Office in Singapore.
The company’s mission is to build photonic switches for AI infrastructure, solving fundamental networking bottlenecks and enabling high bandwidth, low latency connectivity for improved AI model performance. Their optical circuit switch keeps the payload fully optical while directing data to where it needs to go, eliminating the need for expensive transceivers.
Salience Labs’ photonic switch enables all-optical, multi-hop large cluster connectivity, addressing the flexibility and control required in next-generation AI infrastructure. The switch allows for resizing and provisioning networks quickly and on the fly, with fast reconfiguration, novel flow routing, and synchronization technologies.
The company’s technology is the result of over 10 years of research into silicon photonics. Salience Labs’ switch is built for AI and focuses on compatibility, reliability, and manufacturability. The team comprises experts in photonics, electronics, embedded software, and machine learning, working together to shape the future of optical networking.
Salience Labs’ board and advisors include industry leaders such as Daniel Armbrust, Chairman; Dr. Bill Jeffrey, Board Member; Jack Edmondson, Board Member; Jalal Bagherli, Investor and Advisor; and Prof. Wolfram Pernice, Co-Founder and Advisor.
