Swissi AI
Data Centre
Federated AI infrastructure with verifiable storage and ESG integration — Swiss-compliant, sovereign, auditable
A federation of modular data centres distributed across Swiss territory, designed for AI inference, smart contract automation, decentralised storage and ESG-aware orchestration. Coordination through formal game-theoretic incentives within a tokenised and auditable framework.
"Centralised AI infrastructure scales yet conflicts with latency, auditability and energy constraints. The Federated AI Infrastructure offers a policy-aligned alternative for nations that face energy, land or sovereignty constraints."
Formal economic modelling
Unlike conventional infrastructure projects, the Federated AI Infrastructure is grounded in rigorous academic research. Our coordination model applies non-cooperative game theory and formal mechanism design to ensure stable, self-enforcing equilibria.
Each operator optimises availability, effort and energy mix to maximise discounted utility. Under bounded multipliers, corridor constraints and responsive congestion pricing, best responses are monotone and a Nash equilibrium exists.
Stable configuration where no participant can unilaterally improve their outcome given others' strategies
Incentive-compatible structures that align individual optimisation with system-wide objectives
Bounded multipliers and price corridors ensure predictable, stable market dynamics
Resource allocation optimised across computational, energy, monetary and network domains
Hyperscale centralisation
Global data centre electricity demand will more than double by 2030, reaching approximately 945 TWh. Current practices prioritise performance and centralised control over energy efficiency, auditability and regulatory decentralisation.
- Energy crisis: Data centres account for 1-2% of global electricity, with AI workloads as primary driver
- ESG conflict: Microsoft reported 168% increase in AI-related energy demand and 23-29% rise in emissions since 2020
- Sovereignty concerns: Extraterritorial regimes like the US CLOUD Act compromise data localisation
- Auditability gaps: Absence of standardised reporting limits transparency and regulatory oversight
Federated AI Infrastructure
Separates centralised training from decentralised inference and storage across five node classes, coordinated by a verifiable orchestrator.
- Type I: Centralised AI training facility
- Type II: Orchestration and coordination
- Type III: Inference, storage, ledger nodes
Environmental impact expressed through internal, measurable indicators. Carbon-aware scheduling shifts load toward low-carbon intervals.
- Real-time renewable energy tracking
- Carbon-adjusted pricing and settlement
- Smart contract carbon credit module
Dual-token mechanism for compute allocation and energy balancing. Corridor-bounded prices with capped multipliers for SLO, ESG and diversity.
- Posted-price marketplace
- Size-neutral availability floor
- Nash equilibrium coordination
Permissioned DAG with asynchronous Byzantine fault tolerance. All actions cryptographically verifiable, tamper-proof, and audit-ready.
- Leaderless aBFT consensus
- On-chain attestations and metering
- Regulatory proof generation
Sustainable energy by P3 Energy
The Federated AI Infrastructure is powered by P3 Energy, our dedicated energy partner within the Swissi ecosystem. P3 delivers certified renewable energy sourcing, grid integration, and energy-to-compute optimisation for all node classes.
This vertical integration ensures complete supply chain control over energy provenance, enabling verifiable ESG attestations and carbon-adjusted pricing at the infrastructure level.
Learn more about P3 EnergyCertified green energy for all Type III nodes
Live renewable share and carbon intensity
Strategic multipliers for solar, wind, hydro
Verifiable energy provenance per workload
Five standardised node classes
High-throughput inference, ledger validation
Institutional operators
Compute-intensive inference, batch processing
Enterprise partners
Interactive inference, balanced workloads
Regional partners
Edge inference, local storage
Municipalities, cooperatives
Low-latency edge, continuous DAG participation
Individual operators
All Type III nodes integrate the full operational stack: decentralised inference, certified data storage, and distributed ledger operations. Operators can include municipalities, cooperatives, academic institutions and private actors.
Four equilibrium domains
Decentralised workload alignment under rational utility maximisation. No node benefits from unilateral workload reassignment.
Task allocation to nodes under heterogeneous power sources. Energy-adjusted utility balancing throughput, cost and ESG compliance.
Token-based compensation for decentralised task execution. Profit-seeking operators maximise earnings relative to operational costs.
Spatial and temporal feasibility under latency, bandwidth and connectivity constraints. Physical delivery within service-level thresholds.
Federated AI Infrastructure Design
Our research applies non-cooperative game theory and formal mechanism design to coordinate distributed AI infrastructure. We prove that under bounded multipliers and corridor constraints, Nash equilibria exist and best responses are monotone.
The framework addresses four equilibrium domains — computational, energy, monetary, and network — enabling sovereign, ESG-compliant infrastructure that scales without sacrificing auditability or regulatory alignment.
Swissi Authors: Prof. Dr. Walter Kurz, Wojtek Stricker, Prof. Dr. Velimir Dedić
Sovereign AI infrastructure market
Global data centre and AI infrastructure market, with AI workloads as primary growth driver
Sovereign, ESG-compliant infrastructure in Europe, MENA, and Asia-Pacific jurisdictions
Addressable through Swiss deployment and federated partnerships with aligned jurisdictions
Development progress
Full system architecture with five node classes, three node types, and four equilibrium domains
Permissioned consortium DAG with leaderless aBFT consensus and on-chain attestations
Carbon-aware scheduling, renewable tracking, and smart contract carbon module
Type I and Type II facility locations under evaluation with cantonal energy authorities
Tokenised compute marketplace
Revenue through tiered token rewards for verified work, with multipliers for ESG performance, SLO attainment, and network diversity. Token redemption via regulated issuer with segregated reserves.
Swiss-compliant by design
Aligned with Swiss Federal Act on Data Protection and GDPR through privacy by design, data subject rights, and extraterritorial application
Token redemption through regulated issuer with segregated reserves, KYC/AML compliance, and travel rule adherence
Formal measurement of infrastructure performance and energy use for ESG auditability and operational compliance
Projected outcomes
Infrastructure-focused commercial model with long-term recurring revenue through compute marketplace fees, storage services, and cross-jurisdictional licensing.
Capital allocation
- Type I training facility
- Type II orchestration node
- DAG infrastructure
- Initial Type III network
- ESG orchestration layer
- Token marketplace
- Carbon module
- Metering systems
- Type III node deployment
- Partner onboarding
- Regional coverage
- Capacity scaling
- Legal & compliance
- Team expansion
- Energy contracts
- Working capital
Access the full details
Register as a qualified partner to view complete technical specifications, financials, and partnership terms.
Not a partner? Contact us for partnership inquiries