Estonia plans €20 million purchase of AI computing capacity
Tuesday 4th August 2026 on 09:15 in
Estonia
Estonia plans to buy up to €20 million worth of high-performance artificial intelligence computing services from a branch of an AI gigafactory to be built in the country, ERR reported.
The purchase is part of a European Union tender organised by the European High-Performance Computing Joint Undertaking. The project aims to establish up to seven AI gigafactories across Europe, with participating member states and the European Commission planning to acquire services worth up to €10 billion in total, the Ministry of Justice and Digital Affairs said.
Seventeen other EU member states have joined Estonia in signing the agreement for the joint tender. Consortia formed specifically for the tender may include companies, public-sector institutions, investors and other partners.
Estonian companies can form consortia and submit bids directly to the European High-Performance Computing Joint Undertaking through the eSubmission digital system. The successful consortia will later provide participating countries and the European Commission with guaranteed access to the required computing capacity.
The planned gigafactories do not have to be located in a single country. They may also consist of cross-border and distributed AI infrastructure. Through the Ministry of Justice and Digital Affairs, the Estonian government is prepared to purchase up to €20 million worth of high-performance AI computing services between 2028 and 2032 from a consortium that successfully bids to build a new-generation AI data centre branch in Estonia.
Bids must be submitted by November 12, 2026. Contracts are expected to be signed in early 2027, and the first AI data centres should begin operating no later than 18 months after the contracts are signed.
The European Commission says the tender aims to strengthen Europe’s technological sovereignty and give startups, large companies, researchers and the public sector access to infrastructure needed, among other things, to train large AI language models and process data.