Greece has launched Pharos, the country’s first AI Factory, positioning itself within Europe’s early efforts to build large-scale artificial intelligence infrastructure. Announced in December 2024 and formally launched in April 2025, Pharos is part of the European AI Factories programme under the EuroHPC Joint Undertaking. Backed by around €30 million in funding, shared equally between the EU and the Greek government, the project will deliver high-performance computing resources, technical expertise and training opportunities. Its ambition is straightforward: strengthen Greece’s role in Europe’s digital future while tackling applied challenges in health, culture and sustainability.
Key Highlights
Greece in the First Group of EU AI Factories
Pharos is one of seven facilities selected across Europe in the programme’s first round. Being included at this stage gives Greece early influence in shaping the direction of AI infrastructure at continental level, while signalling European trust in its capacity to deliver.
€30 Million Over Three Years
The initiative comes with a budget of roughly €30 million, co-financed by Brussels and Athens. With a three-year horizon, partners are under pressure to show results quickly, from technical delivery to measurable benefits for research and industry.
A Broad Coalition of Partners
The project is led by GRNET under the Ministry of Digital Governance and brings together the Demokritos Research Centre, the National Technical University of Athens, the Athena Research Center and the state investment fund GrowthFund. The consortium blends public oversight with academic and technical expertise, reflecting Europe’s emphasis on collaboration rather than purely commercial models.
Supercomputer DAEDALUS at the Core
Pharos will rely on DAEDALUS, Greece’s recently installed national supercomputer, offering close to 90 petaflops of performance. A substantial share of its processing power will be reserved for AI workloads, enabling projects in data-heavy fields such as natural language processing and medical research.
Clear National Priorities
Unlike some other countries that have emphasised industrial or financial uses, Greece has identified three domains: health and personalised medicine, language and cultural heritage, and sustainability. The focus reflects local needs, from preserving linguistic resources to addressing climate risks in the Mediterranean.
Why This Matters for Power Loop Readers
Energy and Sustainability Challenges
Running a supercomputer of this scale requires significant electricity. Whether Greece can meet this demand through renewable sources or must lean on conventional power will shape the environmental footprint of the project. For the energy sector, Pharos is a reminder that digital and climate strategies are now intertwined.
Skills and Retention
Greece has long faced “brain drain” in the technology sector. Pharos aims to counter this trend by creating advanced research opportunities at home. If successful, it could help anchor a skilled workforce in the country and feed into regional innovation ecosystems.
Europe’s Distributed Model
Spain, Finland and Germany are also setting up AI Factories, each with a different emphasis. The Greek choice of health, culture and sustainability sets it apart and contributes to a broader European mosaic. For observers, this shows how Europe is betting on diversity of expertise rather than a single, centralised model.
Source Listing
- GRNET.gr – Press Release: AI Factories – Greece among the first European countries to host an AI Factory (December 12, 2024) – https://grnet.gr/en/2024/12/12/pr-ai-factories/
- AthenaRC.gr – One of the first seven AI Factories in the EU to be established in Greece – “Pharos” places Greece on the Artificial Intelligence Map (December 12, 2024) – https://www.athenarc.gr/en/news/one-first-seven-ai-factories-eu-be-established-greece-pharos-places-greece-artificial
- EKT.gr – The Greek “Pharos” is among the first seven Artificial Intelligence Factories established in Europe (December 19, 2024) – https://www.ekt.gr/en/news/30578
