Salience Labs is positioning its photonic switching technology to address the escalating latency and performance bottlenecks inherent in modern AI datacenter architectures. The Oxford-based company has secured $125 million in an oversubscribed Series B funding round to accelerate the deployment of its optical circuit switching (OCS) technology. Led by Bullhound Capital, the investment includes participation from major semiconductor players AMD Ventures and Intel Capital, signaling a concentrated industry interest in silicon photonics. This capital injection is intended to scale volume manufacturing operations and expand the company's physical footprint into Silicon Valley, targeting the high-growth market for AI back-end network infrastructure.
Salience Labs $125M Series B Funding and Expansion
The $125 million Series B round, which saw Bullhound Capital’s Callum Stewart join the Salience Labs Board of Directors, marks a significant capital infusion for the company's photonic chip development. Beyond the primary leads, the round drew participation from a diverse group of institutional and strategic investors, including SBI Group, Hudson River Trading, British Business Bank, Applied Ventures LLC, Oxford Science Enterprises, Cambridge Innovation Capital, and Silicon Catalyst Ventures. This financial backing is specifically earmarked to transition Salience Labs from research-heavy development into high-volume production environments.
To support this transition, the company plans to expand its presence in the United States by establishing a second West Coast location in Silicon Valley. On the technical side, the funding aims to advance the company's roadmap beyond its recently demonstrated 64-port all-optical photonic switch, which debuted at ECOC 2026. Salience Labs is now targeting the development of 128-port and 256-port OCS solutions. To achieve the necessary manufacturing scale, the company will continue its engagement with industry partners Keysight Technologies and Tower Semiconductor. This move is designed to ensure that its integrated on-chip solutions can meet the rigorous volume requirements of global datacenter deployments.
Addressing AI Datacenter Scale-Up Bottlenecks
The strategic motivation behind Salience Labs’ architecture lies in the shifting requirements of AI infrastructure, which is increasingly being designed at the rack scale. As AI workloads demand higher token economics, the networking layer becomes a primary determinant of system efficiency and cost. Salience Labs is developing an all-optical, fully integrated switch architecture designed to connect thousands of GPUs across multiple racks. This approach aims to optimize scale-up network performance, providing a high-speed, ultra-low latency fabric that mitigates the compute bottlenecks currently facing large-scale AI clusters.
The market timing for this technology is underscored by projections from the Dell’Oro Group, which forecasts that cumulative spending on datacenter switches for AI back-end networks could approach $1 trillion between 2026 and 2030. Within this projected market, scale-up networking is expected to account for more than half of the total spend by 2030. By utilizing optical circuit switching, Salience Labs is attempting to provide a solution to the power consumption and performance challenges that traditional electrical networking faces when managing massive datacenter capacity. The company’s proprietary optical amplification and integrated silicon photonics architecture are the core components intended to facilitate this transition to all-optical connectivity.
Key Takeaways
- Salience Labs closed an oversubscribed $125 million Series B funding round led by Bullhound Capital, with participation from AMD Ventures and Intel Capital.
- The company plans to expand its manufacturing and commercial presence by opening a second West Coast office in Silicon Valley and scaling production with partners like Tower Semiconductor.
- Following a 64-port switch demonstration, the company is working toward 128-port and 256-port optical circuit switching (OCS) solutions for AI datacenters.
TechInsyte's Take
In our view, the heavy involvement of both AMD Ventures and Intel Capital in this round suggests that the semiconductor industry views photonic switching not merely as an incremental upgrade, but as a fundamental requirement for the next generation of AI hardware. As GPU clusters grow in complexity, the "tax" paid in latency and power when moving data between racks becomes unsustainable for traditional electrical switching. Salience Labs is betting that an all-optical, integrated on-chip approach can bypass these physical limits. However, the success of this strategy depends entirely on their ability to move from successful laboratory demonstrations—like the 64-port switch—to the high-volume, high-yield manufacturing required by hyperscalers. If they can successfully scale to 256-port architectures through partners like Tower Semiconductor, they will be positioned at the center of a trillion-dollar networking shift.
Questions & Answers
How will Salience Labs use the $125 million in new capital?
The company intends to accelerate the deployment of its optical circuit switching (OCS) technology, scale its volume manufacturing operations, and expand its U.S. presence by opening a second West Coast location in Silicon Valley.
What specific technical milestones is the company targeting following its ECOC 2026 demo?
Following the live demonstration of its 64-port all-optical photonic switch, Salience Labs is targeting the development of future 128-port and 256-port OCS solutions to support data-intensive environments.
Why is optical circuit switching becoming critical for AI infrastructure?
As AI infrastructure moves toward rack-scale design, networking directly impacts system performance, efficiency, and economics. All-optical connectivity is positioned to address bottlenecks in performance, power consumption, and cost caused by growing datacenter capacity demands.
Which industry partners are supporting Salience Labs' manufacturing scale?
The company is engaging with Keysight Technologies and Tower Semiconductor to scale its manufacturing operations for high-volume production.
Source: Salience Labs
