How much AI capacity
does Sweden need?

Pull the levers and see how assumptions affect Sweden's total AI compute need — from public sector to the full economy.

Sectors
Presets
62%
25%
55%
80
Sovereign AI training
Training Swedish/EU-owned foundation models
Year
Compute need 2029
6 409
H100 equivalents
~1 423
MSEK/yr
~5.6
MW
~1 020
Gross jobs 2029

A91–A93 via the healthcare adoption slider. Fine-tuning (Tier 3) does not affect this figure. 14-jobb.md →

Compute need over time
Public sector4 850
Healthcare1 559
Energy need 2029
~5.6 MW
Annual cost
~1 423 MSEK

~6 MW = a mid-sized data center. Facebook's Luleå started at ~40 MW.

All assumptions and calculations are open. Help us improve the analysis — especially for sectors with weak data quality. Contribute via PR on GitHub

The public-sector AI question is no longer about isolated pilots. It is about building durable capability where compute, talent, and governance scale together.

The goal is to make trade-offs explicit: what drives compute demand, which risks follow from delayed decisions, and which practical priorities matter most by 2029.

In European context

Sweden's compute need compared to Nordic and European AI investments

LUMI (EuroHPC) (Finland)
5 000
Gefion (NVIDIA) (Danmark)
1 528
Leonardo (EuroHPC) (Italien)
3 500
EU AI Factories (plan) (EU)
50 000
04-internationella-jamforelser.md

Why this matters

1

More than copilots

Agentic workflows, longer contexts, and background agents significantly increase the need. Tier 1 lands around ~2,200 H100-eq in 2029.

03-berakningsmodell.md
2

Sovereignty is a policy choice

Sovereign training accounts for ~4,500 H100-eq — half the base scenario. An active policy decision, not a consequence of user growth.

08-suveranitet.md
3

Current budgets are insufficient

Existing IT budget logic supports ~2,000–4,000 H100-eq. The base scenario requires targeted state investments, EU funding, and public-private partnerships.

03-berakningsmodell.md
4

The full economy needs 4–5× more

Public sector is ~20% of Sweden's total AI compute need. Private sector, research, and defense add up. Energy infrastructure and grid capacity must be planned for the whole picture.

11-kompletterande-perspektiv.md

Why now?

GPU deliveries have 12–18 month lead times. Data centers require grid connections and permits. Every quarter without a decision means a quarter without capacity in 2028–2029.

10-kan-vi-vanta.md

Three recommendations

1

Start procurement now

Framework agreements, vendor dialogue, site selection, and grid connections must start before demand peaks in 2028–2029.

2

Build a hybrid model

Start with ~1,000–1,500 H100-eq: on-prem for sensitive data, cloud for burst.

3

Couple compute with competence

Compute without governance yields low impact. Bundle with accountability, training, and data policy.

Jobs created

AI investments create new roles — not just compute costs

~1 020gross jobs in healthcare AI implementation by 2029 in the base case
Direct roles

Clinical informatics, MLOps, AI safety & compliance, product owners, change management

Indirect roles

System integrators, domain consultants, trainers, independent auditing, vendor support

Gross jobs — net impact depends on how fast administrative tasks are automated and how workforce transition works in practice.

14-jobb.md

All assumptions and calculations are open. Help us improve the analysis — especially for sectors with weak data quality.

Contribute via PR on GitHub

What could go wrong?

What if the money doesn't come?

🏛
Decision-maker

Without targeted investment, capacity stalls at ~2,000–4,000 GPUs — a fraction of what's needed. Sweden falls behind countries investing now.

🏥
Healthcare

Hospitals wanting AI diagnostics become dependent on expensive cloud services — or wait.

👤
Citizen

Processing times stay long. AI tools that could shorten waits in healthcare and government are delayed.

What if AI spreads slower than expected?

🏛
Decision-maker

The compute investment risks standing unused. But it's a more manageable risk — capacity can be leased out.

🏥
Healthcare

AI pilots that work in labs take longer to reach routine care.

👤
Citizen

The change isn't felt yet. But the world around us keeps moving.

What if we can't buy GPUs in time?

🏛
Decision-maker

GPU deliveries have 12–18 month lead times. Every quarter without an order means a quarter without capacity in 2028–2029.

🏥
Healthcare

AI tools exist but can't run locally — sensitive patient data must be sent abroad.

👤
Citizen

Sweden has the technology but not the infrastructure. Like having electric cars but no charging stations.

What if the power grid isn't enough?

🏛
Decision-maker

Data centers need grid connections. Lead times for new connections in Sweden: 2–5 years. That's longer than GPU lead times.

🏥
Healthcare

Regional data centers can't expand without sufficient power capacity.

👤
Citizen

AI infrastructure competes for the same grid as homes and industry. Planning must happen now.