Breaking the energy curve for the AI supercycle
People and Planet impact snapshot
reduction in energy use in Mobile Backhaul from traffic-aware AI optimization by intelligently orchestrating sleep modes
reduction in Massive MIMO energy use during inactive periods compared to active periods using Extreme Deep Sleep mode
reduction in cellular base station energy consumption thanks to network planning and optimization with AI and digital twins
reduction in energy use for FP5 chipset while increasing system throughput by nearly 4x compared to previous generation
When looking at Nokia’s emissions, 94% of Nokia’s total value chain greenhouse gas (GHG) emissions come from the customer use phase i.e.. our customers using our products, which is why enabling customers to decouple traffic from energy consumption is central to our climate strategy.
The challenge is to support rapidly growing network traffic without allowing energy consumption to grow at the same rate. By 2035, Nokia projects that total network traffic will nearly quadruple from 2026 levels to almost 3,789 exabytes (EB) per month, with 44% of that traffic driven by AI. Against this backdrop, GSMA data shows that between 2019 and 2024 mobile data traffic more than quadrupled, while operators’ operational emissions fell by 13%. This makes continued improvements in network energy efficiency critical as networks expand to support the AI era.
Traffic-aware AI optimization uses artificial intelligence to monitor and improve how traffic flows through a system, reducing energy consumption by up to 25%. AI-native Nokia solutions that support energy savings include:
- Extreme Deep Sleep mode in 5G AirScale radios
- AI-optimized radio access networks (RAN)
- Traffic-aware sleep modes
- Energy-efficient site solutions
- Out-of-band management (OOBM)
Extreme Deep Sleep mode in our 5G AirScale radios and AI-optimized RAN are designed to optimize energy consumption and minimize environmental impact. Extreme Deep Sleep mode enables our AirScale MIMO radios to reduce energy use by up to 95% during inactive periods.
Nokia Doksuri: Efficient and lightweight
In 2026, Nokia expanded our AirScale portfolio with Doksuri Radios to advance energy efficiency, capacity, and deployment simplicity. Engineered for real-world performance, Doksuri Radios deliver up to a 30% improvement in power efficiency than previous by combining higher power output with lower energy consumption under typical network conditions. They are also 25% lighter, which lowers sites’ footprints and simplifies deployments.
To support our climate and sustainability objectives, Nokia is also committed to enhancing energy efficiency at our sites. In 2026, Nokia Bell Labs teamed up with KDDI Research to create a more energy-efficient base station through Intelligent 4D Resource Optimization. This new technique helps a base station adjust its radio resources to traffic demand in real time, reducing unnecessary energy use. When testing this technique, we noticed up to 40% lower power consumption at the same throughput levels and four times greater energy efficiency than today’s 5G base stations, while increasing throughput without using more power.
This aligns with our findings that network planning and optimization using AI and digital twins can cut average cellular base station energy consumption by up to 15%.
Similarly, the Aurelis solution for data centers provides a simpler, more efficient, and cost-effective out-of-band management (OOBM) architecture. By reducing the number of active switches by 90%, the solution can improve space efficiency, deliver power savings of 50%-plus, and simplify operational efforts by up to 80%.
Nokia FP5 silicon powers Zayo network
Nokia is pioneering chip design and sustainability. Evolving our chipset from FP4 to FP5 adds more capacity and security capabilities while reducing power consumption. Nokia’s new generation of network processor silicon (FP5) has allowed us to achieve up to 75% reduction in energy use while increasing system throughput by nearly four times.
In 2025, we partnered with U.S. communications infrastructure provider Zayo to deploy a next-generation IP network that’s powered by our advanced FP5 silicon. The first phase was deployed in New York and New Jersey, with the solution intended to enable rapid deployment, increase network reliability, and simplify Zayo’s network operations. Expansion is planned for dozens of additional markets.
Learn more
Blog