Imagine if your computer's power bill and carbon footprint were visible for every single task you ran, and the system could automatically move that task to a 'greener' server. This project builds a smart dashboard and a set of rules to stop data centers from wasting electricity. It's like adding an eco-mode to the giant supercomputers that power modern science.
Digital infrastructures are seeing a massive increase in energy consumption, contributing up to 4% of global greenhouse gas emissions. There is a lack of standardized tools to monitor and optimize this energy use across the entire lifecycle of digital services.
A reference architecture for sustainable infrastructures and a set of tools for energy monitoring, job scheduling, and resource allocation. It also includes a certification framework and training programs for energy-efficient management.
If you are a provider dealing with skyrocketing electricity costs for AI training — this project developed a reference architecture and brokering logic that optimizes task execution toward low-energy use. This helps reduce the environmental footprint of high-intensity workloads.
If you are an operator dealing with strict EU Green Deal reporting requirements — this project developed end-to-end energy monitoring tools and an environmental impact assessment methodology. This allows for precise tracking of greenhouse gas emissions across the digital continuum.
If you are a developer dealing with inefficient code that wastes server resources — this project developed Reproducibility as a Service (RaaS) and energy usage reporting. This enables the creation of energy-conscious applications that maintain experimental accuracy while using less power.
Based on available project data, no pricing or cost information is provided as this is a Horizon-RIA research project.
Yes, the solutions are being validated across four major European digital infrastructures (EGI, SLICES, SoBigData, and EBRAINS) to ensure they are reusable across the broader ESFRI landscape.
Based on available project data, specific licensing terms are not mentioned, though the project focuses on providing building blocks and guidelines for other infrastructures.
The project provides policy recommendations and a certification framework to align digital operations with the European Green Deal and UN Sustainable Development Goals.
Integration occurs through the extension of existing workload managers, Virtual Machine Managers, and AI/ML training frameworks with new brokering logic.
The consortium is heavily weighted toward research and academia, with 13 out of 15 partners being universities or research organizations. Only 1 industry partner and 1 SME are involved, resulting in a low industry ratio of 7%. This suggests the output is currently more focused on technical validation and standards than immediate commercial productization.
Contact the Universiteit van Amsterdam
Contact us to explore how to implement these energy-monitoring building blocks in your data center.
