AI-generated summary
The rapid expansion of the digital world has led to a significant increase in global energy consumption, with data centers and networks accounting for 1% to 1.5% of worldwide electricity use. In response, Europe has introduced initiatives like the Green Cloud & Green Data Centres to align digital growth with climate goals by promoting renewable energy use, enhancing efficiency, and reusing waste heat. The EU Code of Conduct on Data Centre Energy Efficiency provides technical guidelines to reduce electricity consumption through continuous energy monitoring, optimized cooling, clean energy integration, and circular infrastructure design.
Technological advancements are also crucial as AI models grow more complex, making energy efficiency a strategic priority. Innovations such as liquid cooling, thermal immersion, and edge computing reduce energy use and network load. Companies like Equinix and Amazon Web Services lead with transparent energy-saving measures, including advanced cooling systems and energy-efficient chips. Additionally, repurposing waste heat to power district heating systems enhances overall efficiency and offers economic benefits. Europe is progressing toward legislation to enforce data center efficiency and aims for climate neutrality by 2030. The future of digitalization hinges on balancing increased data processing and storage with reduced energy consumption, marking a critical shift toward a sustainable digital and energy revolution.
A tour of data center energy efficiency: metrics, technologies, and the players that are redefining it.
The digital world is growing at a rate that few imagined a few decades ago. Every click, every search, and every model of AI adds to a global infrastructure that consumes as much energy as entire countries. The International Energy Agency It recalls that data centers and networks already account for between 1% and 1.5% of global electricity, and their demand curve continues to rise.
Faced with this reality, Europe has decided to intervene. The European Commission is promoting a framework that aims to make digital growth compatible with the climate transition. The Green Cloud & Green Data Centres initiative proposes that the centres operate with renewable energies, improve their efficiency and reuse the waste heat that is lost today.
In this context, the EU Code of Conduct on Data Centre Energy Efficiency, developed by the Joint Research Centre, has become the technical reference for operators across the continent. Its guidelines establish verifiable practices to reduce electricity consumption: continuous measurement of PUE (Power Usage Effectiveness, standard indicator to measure the energy efficiency of a data center), optimized cooling, integration of clean energies and circular design of the infrastructure.
The challenge, however, is no longer only regulatory, but also technological. As AI models increase in size and complexity, the Energy efficiency ceases to be a technical detail and becomes a strategic factor. Global operators know this: liquid cooling and thermal immersion, which a few years ago seemed like experimental technologies, are beginning to become more widespread.
The American multinational Equinix, specialized in data center management, transparently publishes its approach to energy efficiency, from adopting renewable energy to implementing advanced refrigeration and operating at higher temperatures to reduce HVAC usage. According to the company, the use of systems developed by the Swedish company Munters, allows it to achieve a PUE close to 1.2 in temperate climates.
However, the conversation about Digital sustainability goes beyond how a server is cooled—it includes how data is processed and moved. Edge computing is consolidating itself as a mechanism to reduce latency and consumption: processing closer to where the data is generated avoids unnecessary traffic, alleviates the load on central data centers and decreases the energy use of networks.
Hardware is also experiencing its own green transition. Amazon Web Services offers Clear figures: Its Graviton chip, specifically designed for efficiency, reduces the power consumptionor up to 60% compared to equivalent general-purpose processors. And AI chips, such as Trainium, are optimized to maximize performance per watt in machine learning loads.
The reuse of waste heat is becoming another central element of this new energy logic. In fact, the heat generated by servers, which was previously dissipated unused, can power district heating systems. Recent studies show that their conversion using thermodynamic cycles can significantly improve the overall efficiency of the site and generate economic benefits.
Beyond intentions, Europe is also making progress in regulation. Various legal analyses anticipate that in 2026 the EU will introduce a specific legislative package for data centre efficiency, with the aim of aligning them with data centre commitments. Climate neutrality by 2030.
Everything points to a change of cycle. Digitalization can no longer be understood only in terms of speed and capacity; it should be measured by its energy intensity, environmental impact and circularity. The data economy is entering its decisive decade: that of demonstrating that it is possible to process more, store more and learn more… consuming less. The digital revolution will only be sustainable if it also becomes an energy revolution.
