Ayar Labs announced a partnership with Wiwynn to integrate optical I/O technology into rack-scale AI infrastructure platforms. The companies said the collaboration will combine Ayar Labs’ silicon photonics-based optical interconnect technology with Wiwynn’s rack-scale system design capabilities to enable high-bandwidth, energy-efficient connectivity in next-generation AI data centers.
The companies said the partnership targets emerging AI infrastructure architectures in which optical connectivity moves closer to compute and accelerator devices. Instead of relying solely on electrical interconnects and traditional pluggable optical modules, optical I/O technologies enable direct optical links between processors, accelerators, and memory systems. This approach is designed to reduce power consumption and extend bandwidth scalability as AI clusters grow.
Wiwynn, a major supplier of hyperscale server and rack systems, said integrating optical I/O into rack-scale architectures can help address the bandwidth and power challenges associated with large AI training clusters. Ayar Labs’ technology uses silicon photonics to transmit data using light rather than electrical signals, allowing significantly higher data throughput while reducing energy consumption per bit transmitted.
Key Points
• Ayar Labs and Wiwynn partner to integrate optical I/O into rack-scale AI infrastructure
• Collaboration combines silicon photonics interconnects with hyperscale rack design
• Optical I/O replaces traditional electrical SerDes links between chips and systems
• Technology designed to increase bandwidth and reduce energy consumption in AI clusters
• Partnership targets next-generation AI data center architectures
“Ayar Labs’ optical I/O technology enables the bandwidth density and energy efficiency required for next-generation AI infrastructure,” said Mark Wade, CEO and co-founder of Ayar Labs. “Working with Wiwynn allows us to integrate optical connectivity into rack-scale systems that power the world’s largest AI deployments.”
