IP Infusion has announced OcNOS 7.0, a network operating system release it says is designed to improve AI data center fabric efficiency and simplify the move to 400G and 800G open converged IP and optical transport using IP over dense wavelength-division multiplexing (IPoDWDM). IP Infusion introduced the release at Mobile World Congress (MWC) 2026 in Barcelona and says it will demo IPoDWDM scaling and “AI-ready” networking operations at the event.
For AI data centers, IP Infusion says OcNOS 7.0 targets low-latency, lossless networking for GPU back-end fabrics, plus “open automation” to speed deployments and support “secure operations at scale.” The company also claims that its commercial open model on white-box hardware can deliver 40% to 60% total cost of ownership (TCO) savings versus closed, vertically integrated platforms.
On the transport side, IP Infusion positions OcNOS 7.0 as an IPoDWDM platform that supports open hardware platforms and optics from multiple vendors, including native 400G and 800G ZR and ZR+ optics. The release also includes a full Segment Routing feature set, adding Segment Routing Traffic Engineering, Border Gateway Protocol (BGP) Free Core, and Segment Routing with Flexible Algorithm (Flex-Algo). IP Infusion says this is intended to simplify the control plane, reduce label-switching complexity, and enable deterministic, service-aware paths end-to-end across converged IP and optical networks.
For security at the internet edge, OcNOS 7.0 adds BGP Resource Public Key Infrastructure (RPKI)-based invalid-route rejection and hardened BGP peering controls, which IP Infusion says are aimed at reducing exposure to route leaks and prefix hijacking and improving routing-table integrity. For operations, the release adds Docker and Kubernetes support so operators can run monitoring tools, security probes, or custom microservices directly on the router in isolated containers.
OcNOS 7.0 also expands real-time automation and visibility, adding more than 100 new gRPC Network Management Interface (gNMI)-based streaming telemetry sensors, plus on-change streaming and full NetConf and gNMI integration. IP Infusion says these additions are intended to support real-time fabric optimization, faster rerouting, and proactive capacity planning across converged IP and optical networks, including data center interconnect (DCI) deployments.
Source: IP Infusion
