Open Line Systems Are Winning — Slowly, Reluctantly, Inevitably

Every major carrier I spoke with in 2020 told me Open Line Systems was a niche play. Maybe useful for greenfield builds, maybe a defensive negotiating tool, maybe interesting for hyperscalers but not a serious option for tier-one carrier transport. The narrative was consistent. The deployments told a different story.

By 2025, every one of those same carriers had at least one OLS pilot in production. Most had two or three. A handful had OLS as the planned standard for new metro builds. None of them announced the policy shift publicly. The narrative caught up to the deployments about three years late.

OLS is winning. The reason it's winning slowly is operational, not technical. The reason it's winning reluctantly is political. The reason it's winning inevitably is economic.

What an Open Line System actually is

A traditional DWDM transport platform is delivered as a unified vendor stack. Line system (amplifiers, multiplexers, optical channel monitors, ROADMs), transponders (the bits that put bits on wavelengths), and management plane all come from the same vendor. The system is closed in the sense that you can't take a transponder from vendor A and use it on the line system of vendor B without specific interop testing that the vendors typically don't do.

An Open Line System decouples the line system from the transponders. The line system is "open" — it accepts whatever transponder can drive a compliant wavelength. The 400ZR and 400ZR+ standards from the OIF made this technically feasible. You can now buy a line system from vendor A, plug in 400ZR pluggables from a different vendor's switch, and the link works.

This is disaggregation in the same sense that white-box switching disaggregated the IP routing market — a long political fight that the technology eventually settles.

Where OLS won the easy ones first

The hyperscaler intra-DC fabric. Easy. The transponder is a pluggable 400ZR in a switch, the line system is an amplifier-and-OADM chain the hyperscaler buys directly. No carrier in the picture, no vendor lock to argue about. By 2023, all the major hyperscalers had OLS in production for inter-DC connectivity. By 2024, this segment was effectively standard.

Metro carrier networks where the operator was rebuilding the transport plane from scratch. Greenfield builds let the OLS architecture be the design decision rather than a migration project. Several European carriers chose OLS for their 5G transport backhaul refresh because it gave them transponder flexibility and a single line-system stack across multiple wavelength applications.

Regional ISPs serving data-center interconnect markets. These operators had less legacy and more competitive pressure on capex. OLS gave them a cost structure that closed-stack vendors couldn't match.

Where OLS is winning slowly

Tier-1 carrier long-haul. The hard one. Long-haul transport platforms are 10-to-15-year-life infrastructure. Replacing them is risky, expensive, and operationally disruptive. The case for OLS is real but the migration is the problem. Most large carriers have OLS pilots running in parallel to existing closed-stack networks, but the migration to OLS-as-primary is happening on multi-year timelines.

Carrier metro networks with existing legacy. Same problem at smaller scale. The technical case is good, the migration story is hard.

Telco edge networks. The customer-facing access nodes are slowest to change because the operational tooling is built around specific vendor stacks. Replacing the access-side tooling is more disruptive than replacing the transport-side stack.

Why it's still slow despite winning

The technology won the argument. The slowdown is now entirely operational and political.

Operational complexity. A unified vendor stack means one support contract, one set of tools, one management plane. OLS means juggling at least two vendor relationships and writing your own correlation across their telemetry. Most carriers' OSS systems were not designed for this. The migration cost includes rewriting the OSS, not just the transport hardware.

Yellow-wire problem. Faults at the boundary between line system and transponder are operationally the hardest. Each vendor blames the other. The carrier ends up running the integration testing themselves. This is the single biggest reason carrier ops teams resist OLS — it shifts blame to them when something fails.

Procurement contracts. A 10-year transport platform contract with a single vendor includes service-level guarantees that span line system and transponders. OLS deals require unbundling those guarantees, which neither vendor wants to do, which extends procurement timelines.

Internal politics. Vendor sales teams sell against OLS internally to the carrier operations and procurement organisations. The line they take is "OLS adds operational complexity that we eliminate". Carrier middle management often agrees, because their accountability is to the existing operational model.

Where this lands in the next 24 months

Three things I expect by mid-2028.

At least three top-15 carriers will have OLS as their default for new metro builds. Not pilots. Production policy. The first one is probably announced in 2026, with the others following within a year.

Vendor positioning shifts. The vendors who shipped closed-stack platforms will all have OLS-compatible line-system SKUs by 2027 because their customers will demand it. The interesting question is which vendors lead the shift versus which get forced into it.

One major vendor will exit the closed-stack market entirely. The economics for a sub-scale vendor competing in both modes are bad. Expect a consolidation event in 2027 where the smallest of the closed-stack incumbents pivots fully to OLS line systems and gives up the transponder business.

What "winning" actually looks like

OLS is the future of optical transport in the same way disaggregated routing was the future of the data center. The question was never "is this technology right". The question was always "how long does the political and operational drag take to clear".

The drag is clearing. The deployments are real, the carrier policy shifts are happening quietly, the procurement teams are running side-by-side comparisons and they're not coming back. The closed-stack vendors will continue to ship product for a long time, but the strategic momentum has flipped.

OLS won the argument three years ago. The narrative is still catching up. The deployments are what's actually true.