CeADAR Spearheads a Major €8.7 Million EU Project Focused on Energy-Efficient AI for Data Centres and IoT

CeADAR, Ireland's Center for AI, has taken the helm of a significant €8.7 million European Union initiative aimed at enhancing the energy efficiency and trustworthiness of artificial intelligence systems. This groundbreaking project, named Manolo, involves a collaboration of 18 partner organizations across eight European nations. The central objective is to advance AI tools, algorithms, and frameworks that will not only reduce the energy, memory, and storage demands for training and operating AI models but also foster a greater level of trust among users towards AI technology.

Addressing the Trust and Efficiency Challenges in AI

The ambitious Manolo project sets out to tackle two pressing challenges in the AI domain—energy consumption and public skepticism. As concerns regarding the substantial energy requirements of AI operations in data centres and personal devices grow, researchers are driven to develop innovative methods for constructing more efficient AI models. Such advancements could significantly lessen the strain on energy grids due to the high demands of data centers and other AI-driven systems.

In parallel, there exists an observable decline in user confidence in AI technologies. Manolo's project addresses this by striving to establish a framework that reverses this trend. By developing a handbook of practical guidelines, the project seeks to assist companies in building AI solutions that are deemed trustworthy.

Advancing Compliance with EU AI Legislation

The outcomes of the Manolo project are expected to align well with the EU’s proposed AI legislation, the EU AI Act. Companies will be able to use the project's trustworthy algorithms and frameworks to ensure they meet these new regulatory standards. The tools created will undergo rigorous testing in laboratory environments before being validated across various industrial settings, including manufacturing, healthcare, sleep monitoring, and even in mobile phones.

Another intriguing aspect of the project is its experimentation with neuromorphic chips. These processors, inspired by human neural systems, promise to deliver faster performance with reduced power usage compared to conventional computer processors.

Funding and Project Duration

Manolo is supported by Horizon Europe, the EU’s principal funding program for research and innovation, which boasts a hefty budget of €95.5 billion. The program aspires to tackle climate change, aid in attaining the UN's Sustainable Development Goals, and enhance the EU's competitiveness and growth.

The project is scheduled to span over three years, where the development and implementation of these novel AI resources will be carefully monitored and adapted for the best possible integration and efficiency.

The project's coordinator from CeADAR, Dr. Ricardo Simon Carbajo, expressed pride in leading such a substantial EU-funded project, which seeks to create innovative algorithms for more trusting and efficient AI. He underscored the importance of these solutions in enabling organizations within Ireland and the rest of Europe to attain their desired energy efficiency goals while bolstering the reliability of AI applications.

AI, energy, trustworthiness