Home/ What it is used for
Who actually
needs this?
Nobody buys computing capacity for its own sake. Hospitals, weather services, factories and banks buy it for a specific job. Here is what those jobs look like, and why the location of the machine keeps coming up.
01Start with something ordinary
Five ordinary things. Five different reasons the machine has to be here.
Pick one. The three bars show what it asks of the machine, and they change from one to the next. That is the whole argument.
One swipe, one fresh prediction, made before the screen finishes moving.
Tallinn to Helsinki is stated at under 3 milliseconds.[1]
It works today because somebody else built the building. The rules and the money live there too.
It answers from your own contracts and email. Nothing leaves the building.
Public models are ruled out where confidentiality is required.[13]
The choice is two. Send the contracts abroad, or have no assistant.
Text goes in, Estonian comes out, faster than it is spoken.
Estonian language models are built in Tartu and at TalTech.[13]
No foreign company has a reason to do this for a million people.
The model marks what the radiologist should look at first.
An AI rack draws 120 to 130 kW, an ordinary one 8 to 12.[12]
Without a machine under Estonian rules, the hospital simply waits.
Thousands of chips, wired as one machine, running for weeks.
Europe calls the shortage of this capacity a bottleneck.[14]
You use somebody else's forecast, drawn for somebody else's coastline.
The bars compare the five with each other. A rough profile, not a measurement.
02Where it gets used
The European AI factories were set up around specific sectors.
Health
Reading scans and supporting diagnosis. One selected EU site is dedicated to health, and patient data is exactly what cannot easily leave the country.[7][15]
Weather and climate
Listed as a high impact use case in the EU network. Short range forecasts matter for farming, shipping, energy trading and storm warnings.[15]
Manufacturing
Spotting defects on the line and knowing a machine will fail before it stops. Manufacturing and automotive are named focus areas.[15]
Finance
Detecting fraud in payment streams and assessing risk, under tight rules about where the data may sit.[7][15]
Public services
Translation, document handling and citizen support. The EU network includes public bodies alongside startups and researchers.[7]
Startups and small firms
Buying a cluster of AI chips outright is beyond most small firms. Shared capacity is how they get access.[7]
03The Estonian case
A language with a million speakers has to look after itself.
Large models are trained mostly on English. They produce Estonian by translating rather than reasoning in it, which loses the way things are actually said.
TartuNLP and TalTechNLP build Estonian models directly, funded by the Ministry of Education and Research. There is a second reason beyond quality: public foreign models cannot be used where confidentiality is required, such as defence or healthcare.[13]
Associate Professor Kairit Sirts of the University of Tartu puts it plainly: the Estonian language and cultural background does not matter to technology companies, so Estonians have to look after it themselves.[13]
A model needs somewhere to run. That is the link between a research group in Tartu and a building full of power and cooling in Hüüru.
04Why location keeps coming up
The rules that apply to your data depend on where the machine is.
Send patient records to a computer in another jurisdiction and you take on a problem that has nothing to do with technology. Which courts decide disputes. Which government can compel access. Which regulator audits the arrangement.
Keeping the machine inside the EU removes most of that question. This is why governments now fund buildings, not only software.
Estonia signed the EuroHPC agreement but has not been selected as a host site. Lithuania has, through the LitAI Factory.[7][8] Anyone calling Hüüru the first AI factory in the Baltics should be asked which definition they mean. A separate, larger EuroHPC AI Gigafactories tender opened at the end of July 2026, and Estonia has signalled interest in a share of it, with bids due in November. Nothing is decided yet.[21]
05The fair objection
These buildings use a lot of electricity. That objection is legitimate.
The International Energy Agency put data centre electricity use at 415 terawatt hours in 2024, about 1.5% of world consumption, and expects around 945 by 2030, slightly more than Japan uses today. Demand from AI optimised sites alone should more than quadruple.[11]
What honest answers look like
Three levers, all measurable rather than rhetorical: where the electricity comes from, how efficiently the building uses it, and whether the leftover heat is captured.
On the first two, the Hüüru operator publishes 100% renewable supply and a target PUE under 1.2.[3] On the third we found no public evidence either way, so it stays on the unknown list.
Capacity opens
in Q4 2026.
The waitlist opens in the fourth quarter of 2026. Places go in the order they arrive.