AI puts high demands on optical infrastructure, from the physical links to the control plane and everything in between. At Nokia, we help operators build optical networks for AI that deliver on all fronts.
Massive DCI capacity, engineered for performance
Our optical transport solutions are powered by our industry-leading
(PSE) and Infinite Capacity Engine (ICE) coherent optics, which are designed for high-bandwidth, low-latency interconnects. The
Nokia 1830 Photonic Service Switch
(PSS) and 1830 Global Express (GX) solutions support scalable C+L band transmission and provide the power efficiency, reach and spectral flexibility you need for metro, regional and long-haul data center interconnect (DCI) transport.
These capabilities are critical for AI-enabled applications and their workloads, which often require bursty, low-latency transfers across unpredictable routes. Our
and
ensure robust performance, even under extreme traffic conditions and across complex multi-site GPU cluster topologies.
Built-in quantum-safe security
Our optical transport solutions secure AI data sets against emerging quantum threats by integrating at-speed quantum-safe encryption into the network layer using advanced symmetric cryptography and high-entropy key sources. We augment these capabilities with a defense-in-depth approach that provides comprehensive crypto-resilience and ensures all communication layers are quantum-safe.
With our
approach, you can prepare for the post-quantum world today. You get multilayer security that protects sensitive data against cyber agents seeking to harvest this information and decrypt it when cryptographically relevant quantum computers become available—an unknown future date known as Q-Day.
Resilience and assurance at scale
High-performance optics can’t handle dynamic and distributed AI workloads on their own. We source telemetry key performance indicators (KPIs) directly from the network to inform automated assurance processes and enable you to maintain real-time visibility and control across the optical domain.
This telemetry, along with classical AI and GenAI capabilities, lays the foundation for autonomous optical networks that can
. If an optical link degrades or fails, the network can predict and identify issues, assess their impact and reroute traffic without human intervention. This is an important step towards closed-loop automation, which is a precursor to autonomous networking.