Ayar Labs Integrates Co-Packaged Optics with NVIDIA NVLink Fusion for AI Infrastructure

  • Ayar Labs joined the NVIDIA NVLink Fusion ecosystem on June 3, 2026, making its co-packaged optics (CPO) solutions compatible with NVIDIA’s optical and SerDes technologies.
  • The collaboration aims to address bottlenecks in AI infrastructure scaling, particularly in data movement and power efficiency.
  • Ayar Labs’ CPO technology is designed to enable flexible system architectures and efficient AI scaling beyond the limits of copper-based connectivity.
  • The partnership follows Ayar Labs’ $500M Series E funding round, which included participation from NVIDIA.

As AI factories scale to larger GPU counts and increasingly heterogeneous architectures, data movement and power are becoming critical bottlenecks. Ayar Labs’ integration with NVIDIA NVLink Fusion addresses these challenges by providing high-bandwidth, low-latency, and power-efficient connectivity solutions. This partnership is part of a broader trend toward co-designing AI infrastructure to optimize performance and efficiency at scale.

Adoption Pace
The pace at which hyperscalers and system innovators integrate Ayar Labs’ CPO solutions into NVIDIA NVLink Fusion architectures.
Heterogeneous Compute
How the collaboration will impact the deployment of heterogeneous compute within NVIDIA’s rack-scale platforms.
Market Differentiation
Whether Ayar Labs can sustain its competitive edge in co-packaged optics as the NVLink Fusion ecosystem expands.