India-based data center operator CtrlS Datacenters is deploying Ciena optical networking technology across five metro networks that will connect nine data centers in Chennai, Hyderabad, Kolkata, New Delhi and Noida.

The companies said in a September 28 announcement that the networks are designed to deliver up to 100 terabits per second of capacity. The project forms part of a wider Managed Optical Fiber Network deployment for an unnamed hyperscale customer.

Five metro networks link nine facilities

CtrlS will use Ciena’s WaveLogic 5 Extreme-powered Waveserver platform and 6500 Reconfigurable Line System for the rollout. Ciena’s services team is handling end-to-end project management, while its Navigator Network Control Suite is intended to automate network operations and speed service delivery.

The announcement describes a data center interconnect deployment rather than the construction of new facilities. It does not disclose the customer, contract value, rollout schedule or how the planned capacity will be allocated among the five metropolitan areas.

CtrlS founder and chief executive Sridhar Pinnapureddy said the company is seeing demand from enterprises and hyperscalers for more reliable data center interconnect services. Ciena India vice president and general manager Prashant Ramesh Malkani said the upgrade is intended to improve network performance as artificial intelligence workloads add to India’s digital infrastructure needs.

The companies did not publish independent performance measurements for the system. The figure of up to 100 Tb/s should therefore be read as the announced capacity of the deployment, rather than current traffic carried across the network.

Optical links support AI and cloud workloads

Data center operators use metro optical networks to move large volumes of data between facilities with high capacity and low latency. That role becomes more important as cloud platforms and AI clusters spread compute, storage and networking equipment across multiple sites.

The CtrlS project covers major data center markets in northern, eastern and southern India. Connecting facilities in five cities gives the operator a broader platform for serving customers that need resilient links between campuses and metro locations.

The release describes five metro networks, not a single national backbone. It does not say whether the five city networks will be directly interconnected, whether additional long-haul routes are included, or whether the 100 Tb/s figure is an aggregate design capacity or applies to particular links. Those architectural details would be needed to compare the deployment with other large interconnection projects.

CtrlS says it operates 19 data centers across nine Indian markets with more than 370 megawatts of capacity. It also reports 4.4 gigawatts of projects at different stages of execution. Those portfolio and pipeline figures are company-supplied and were not independently verified for this report.

The operator has also announced plans to expand into the Middle East and Southeast Asia, with Thailand identified as its first overseas market. The current Ciena deployment, however, is focused on CtrlS’ domestic Indian footprint.

Interconnection is part of the infrastructure constraint

The rollout illustrates that AI infrastructure expansion involves more than adding servers or building data center capacity. High-capacity links between facilities are also required to move training data, application traffic and cloud workloads reliably across a distributed footprint.

That networking layer complements the power, cooling and site-capacity issues discussed in a recent TNGlobal analysis of India’s AI infrastructure challenge. The comparison is contextual: neither CtrlS nor Ciena disclosed the energy use, customer workload or commercial economics of this optical deployment.

The companies also did not say when all five metro networks will enter service. The next material milestones would be completion dates, customer activation and measured capacity or reliability data once the system is operating.

India’s AI future depends on an infrastructure challenge few are talking about