Skip to main content

Wind River says O-RAN plugfest test cut energy use up to 12%

Wind River said its Fall 2026 O-RAN Global PlugFest demo cut energy usage up to 12% at an Open RAN site.

Wind River says O-RAN plugfest test cut energy use up to 12%

Wind River said its Fall 2026 O-RAN Global PlugFest blog post showed closed-loop energy optimization for an Open RAN site, with software energy profiles cutting energy usage by up to 12%.

The company said the demonstration covered a mobile operator customer’s site, used live O-RAN traffic patterns, and ran on real-world hardware rather than in a lab.

Test setup and traffic patterns

Wind River said it recreated traffic patterns from existing telecom systems and trained an AI model to manage them. The proof of concept used two software energy profiles at the CaaS layer: a low-energy profile with the deep CPU C6 sleep state enabled, and a high-energy profile with C0, C1, and C1E states enabled.

The company said the work was done for an Open RAN site at the Fall 2026 O-RAN Global PlugFest and that the site used live O-RAN traffic patterns. Wind River also said the first three months of testing produced no service disruption and that the AI-based system was still running without human intervention.

Measured energy results

Wind River said the low-energy profile saved 12% with low traffic and 8% with normal traffic. It said the high-energy profile saved 4% with high traffic.

The company also said that using the low-energy profile under high traffic increased energy use by 6%, while switching to the high-energy profile turned that into a 4% savings.

Wind River’s automation model

Wind River said the system used analytics to identify traffic and usage trends, then applied a policy-defined energy profile through orchestration. The company framed the setup as a closed loop: analytics produced the signal, guardrails set the limits, and orchestration applied the action.

The blog post said the model is intended to keep the infrastructure layer from changing at the wrong moment and affecting service.

Droam Perspective

For mobile network operators and RAN teams, the disclosed value is not the PlugFest stage name but the control loop Wind River says it tested against live traffic. The key point is that the company is tying AI-guided energy management to a specific operational constraint: no service disruption during the first three months of testing.

The numbers are uneven by design. Wind River’s own figures show that the same profile can help or hurt depending on load, which makes workload awareness the real product claim here, not a blanket power-saving promise.

Sources

Frequently asked questions

How much energy did Wind River say the test saved?
Wind River said its software energy profiles cut energy usage by up to 12% versus a baseline server without a software energy profile. It also said the low-energy profile saved 12% with low traffic and 8% with normal traffic.
Did Wind River say the site stayed up during testing?
Yes. Wind River said the site experienced no service disruption during the first three months of testing, and that the AI-based system was still running without human intervention.
What profiles did Wind River test?
Wind River said it used two software energy profiles at the CaaS layer: a low-energy profile with the deep CPU C6 sleep state enabled, and a high-energy profile with C0, C1, and C1E states enabled.

DROAM News is published by Droam BV, which sells connectivity services. Droam may be mentioned in or benefit from topics we cover. See our Editorial Policy.