Digital demand is climbing, and energy constraints are quickly becoming one of the biggest pressures on enterprises and network operators. Resilience now depends on infrastructure that conserves as intelligently as it connects — a shift driven by rising OPEX exposure and the ICT sector’s growing contribution to global emissions.

Data centres and communication networks already consume 2–3% of global electricity and account for around 1% of global greenhouse gas emissions, a share expected to grow sharply. Traffic growth is outpacing grid readiness in key technology hubs, and networks are becoming a major part of the expanding load as every increment of data must be transported, processed, and stored. ¹

This is unfolding against a backdrop of accelerating investment: global IT spending in the power and utilities sector is forecast to reach US$249.1B in 2025, rising to US$385.6B by 2029.²

As 5G scales, energy becomes a core operational variable. Radio Access Network (RAN) drives more than 80% of mobile network power consumption, with core networks, cloud infrastructure, and support systems accounting for the remaining 20%.³ This concentration makes the RAN the focal point for next-generation optimisation efforts.

Energy-intelligent networks deliver greater resilience, cost stability, and reduced carbon exposure — outcomes that matter as energy volatility grows. Singtel’s Science Based Targets initiative (SBTi)-approved target to cut absolute Scope 1 and 2 emissions by 55% by 2030 reflects this shift, with network energy optimisation positioned as a central pillar of its long-term resilience strategy.⁴

Traditional networks run at fixed performance levels regardless of traffic. Singtel 5G+ introduces a more adaptive architecture, using AI-driven controls to align network resources with real-time demand. Capabilities such as network slicing improve service consistency by allocating bandwidth, latency, and compute to the applications that need them, while the underlying system can scale activity up or down based on utilisation. The network evolves from a static, always-on model into one that operates with greater precision, efficiency, and responsiveness across varying workloads.

One of the most practical levers for reducing network energy use is allowing components to enter deep-sleep states during low-demand periods without affecting user experience. Singtel has operationalised this through two capabilities: NR Booster Carrier Sleep, which powers down 5G AIR units when they are not required, and RAN Compute Deep Sleep, which lowers consumption from processing elements during quieter traffic windows. Combined, these features deliver an estimated 1,069.60 MWh of annual energy savings.⁴

To put that scale into context: A typical household uses 1–2 kWh per hour⁵, which means 1 MWh (1,000 kWh) can power 500–1,000 homes for an hour.

An average home consumes roughly 10,500 kWh per year, so 1 MWh is enough to power around 300 homes for a day. Applying that to Singtel’s savings: 1,069.60 MWh = 1,069,600 kWh, which is enough to supply approximately 102 homes for a full year.

This quantifies the impact of targeted optimisation and shows how engineering decisions translate into measurable operational and sustainability gains.

Energy use has become a strategic operating variable for modern networks. Singtel’s approach integrates long-term carbon reduction targets with engineering choices that improve how the network allocates and uses resources. The effect is a more reliable and demand-aware infrastructure that reduces unnecessary consumption.

Turn smarter energy use into your next competitive advantage.