Machinepower Index / Dossiers / Finland
Rank 16–22 across the scoring rules tested. How this is tested
Finland
Compute Rentier — For nations whose main asset is the ground itself, meaning cheap clean power, land and stability: host other nations' compute and turn geography into rent, jobs and relevance.
How the Score, Potential and bottleneck cost are worked out: methodology.
The three layers
Energy, grids, data centres and silicon — the physical capacity to run models at scale.
Compute tracked by Epoch AI: Finland hosts 60,000 H100-equivalents for foreign firms. Hosted compute is shown, not scored.
Models, alignment capability, standards influence and the talent that produces them.
The capacity to wire AI into the state and the economy while holding public consent.
Where the score comes from
| Layer | Sub-dimension | Basis | Score |
|---|---|---|---|
| Watts | Planning & Permitting | Assessed | 6.4 |
| Firm Power Capacity | Dataset | 3.8 | |
| Sovereign Compute | Dataset | 6.6 | |
| Hardware & Silicon Access | Dataset | 5.2 | |
| Weights | Frontier Model Output | Assessed | 3.4 |
| Alignment Capability | Assessed | 6.2 | |
| Regulatory Standing | Dataset | 8.4 | |
| Talent Density | Dataset | 4.7 | |
| Will | Policy Capacity | Dataset | 8.1 |
| Public Sector Adoption | Dataset | 8.2 | |
| Workforce Transition | Dataset | 8.3 | |
| Public Trust | Assessed | 4.7 |
Assessed marks a measurement we judged rather than took from a dataset: 4 of the 12 here.
Sources
Agreed connections would raise national demand by two-fifths, so from 2027 large data centres queue behind other users unless they bring their own power.
View source →One of three EU states with an AI Act law in force; no safety statements or safety body.
View source →Largest: LUMI.
View source →Poro 34B, trained from scratch on LUMI on one trillion tokens; its maker, Silo AI, was bought by AMD in 2024.
View source →Talent is a log scale on the count.
View source →Feeds governance and public-sector adoption; policy capacity is listed separately.
View source →Ipsos does not cover it. Mapped to the Ipsos scale through the 28 countries both surveys cover.
View source →Feeds workforce, rescaled across 173 economies.
View source →Fossil, nuclear, hydro and other dispatchable plant; averaged with generation for firm power.
View source →Second source for talent, averaged with Stanford's count.
View source →Bottlenecks
These hold Finland below its Potential. Its two lowest scores cost it the most: frontier model output at 3.4 (part of Weights) and firm power capacity at 3.8 (part of Watts).
Assessment
Finland hosts LUMI, one of Europe's five largest supercomputers, and big foreign campuses, yet controls little of the compute on its soil. Its scratch-trained model, Poro, came from a lab now owned by AMD.
The art of the possible
From 2027 big data centres join the grid behind other users unless they bring their own generation. If Finland uses that to favour campuses that also serve its research, hosting becomes capability.
Strategic play — Compute Rentier
Price grid access in capacity, not cash: campuses that want a connection should share compute with Finnish research and firms.
Leverage — the nearest position. No chokepoint. A stable grid with clean power makes Finland a host Europe trusts; that trust is the asset.
Open Finland in the live Index → — every source behind every number, and the comparison tool.