Oriole, AMD to advance photonic AI networking
Author: Joe Peck
Oriole Networks, a London-based photonic networking startup, has announced further progress in its collaboration with AMD, an American multinational semiconductor company, as part of the UK’s Advanced Research and Invention Agency (ARIA) Scaling Inference Lab programme, including plans to deploy what the company describes as the “world’s first” large-scale AI system based on a pure photonic network.
The project combines Oriole’s PRISM photonic networking technology with AMD Instinct GPUs and AMD EPYC CPUs to explore new approaches to AI infrastructure that aim to reduce latency, improve performance, and lower energy consumption.
According to the companies, the collaboration has been underway for more than a year and is focused on addressing networking challenges associated with increasingly large AI deployments.
Oriole’s PRISM platform replaces traditional electronic switches within the network core with optical circuit-switching technology, enabling data to be transmitted using photons rather than electrical signals.
The company says this approach is intended to reduce network power consumption and minimise latency between computing resources, helping to improve the efficiency of AI inference workloads.
AMD is providing processor and accelerator hardware for the project, alongside technical support to develop and evaluate large-scale AI networking models.
James Regan, CEO of Oriole, notes, “A year ago, we were proving the physics. Today, we’re proving the business.
“Our collaboration with AMD has moved from concept to deployment to a system an order of magnitude larger, and the data proves this is already driving performance increases at pace.
“This is what it looks like when photonic networking stops being a research curiosity and starts being the foundation of how serious AI infrastructure gets built.”
Exploring alternatives to traditional networking
Oriole says PRISM has been designed to operate independently of specific processor or accelerator vendors, allowing it to be deployed across different AI hardware platforms.
The company states that the technology can reduce the reliance on conventional electronic networking equipment while also lowering cooling requirements and associated water consumption.
Madhu Rangarajan, Corporate Vice President, Compute and Enterprise AI Business at AMD, says, “AMD is excited to collaborate with Oriole on the ARIA Scaling Inference Lab cluster.
“Oriole’s AI backend networking with nanosecond optical circuit switching represents a fundamentally different way to connect accelerators at scale. We are helping to validate how photonic fabrics can work alongside AMD compute to deliver the low-latency, high-bandwidth connectivity that AI inference workloads demand.”
The deployment also represents the first commercial implementation of Oriole’s technology, which the company says has progressed from research and development to production readiness within three years.
Suraj Bramhavar, Programme Director at ARIA, comments, “Meeting the demands for modern AI requires rapidly identifying ways to improve the performance and cost-efficiency of large-scale AI clusters.
“ARIA is thrilled to collaborate with Oriole and AMD to demonstrate the benefits of this new technology, and it’s exactly the type of collaboration, between innovative startups and industry leaders, that the Scaling Inference Lab was designed to foster.”
Oriole says wider deployment of its photonic networking technology is planned from 2027 as demand grows for infrastructure capable of supporting large-scale AI workloads.
