Zug, Switzerland — 28 January 2026 — As demand for high-performance compute accelerates across Europe, the question is no longer whether infrastructure will be built, but how it will be structured. The Riveon Group has settled on an answer: Managed Compute Services, delivered on a business-to-business basis through a model that cleanly separates hardware ownership, site operations, and end-use.
At the centre of this architecture is GreenGridLabs (GGL) Finland Oy, the local operating entity at the group’s Dalsbruk datacenter in southern Finland. GGL provides compute as a service to B2B clients — not as a reseller, not as a hosting provider, but as a managed operator responsible for dispatch, scheduling, monitoring, and maintenance of compute workloads across the site’s infrastructure.
The CLU: A Universal Unit of Compute
Capacity is measured in Compute Load Units (CLU), where 1 CLU equals 1 kWh of compute load delivered. This standardised metric allows clients to contract for guaranteed compute capacity in terms that are transparent, auditable, and independent of any specific hardware configuration or downstream workload type. Whether the compute serves AI inference, blockchain validation, or scientific simulation is immaterial to the service contract — what matters is the guaranteed availability of measured capacity.
Flat Fee, Clean Economics
Clients pay a flat fiat-denominated service fee per CLU. Revenue to the operator is deterministic and independent of the results, yields, or outputs of whatever workload the customer elects to run. This structure eliminates the volatility that plagues revenue models tied to commodity prices, token values, or algorithmic performance. For the operator, it means predictable cash flows. For the customer, it means predictable costs.
Three Roles, Clearly Defined
The model rests on a deliberate separation of responsibilities. Riveon AG, domiciled in Switzerland, structures, finances, and governs the hardware — it is the owner of the compute, battery storage, and AI equipment deployed across the group’s sites. GGL Finland, as the local operating entity, manages the physical site: power delivery, cooling, dispatch scheduling, maintenance, and regulatory compliance with Finnish authorities. The B2B customer contracts for compute capacity and determines how that capacity is used.
This tripartite structure is not incidental. It is designed to ensure that each entity bears only the risks and responsibilities appropriate to its role. The owner does not operate. The operator does not speculate on workload outcomes. The customer receives guaranteed capacity without the capital expenditure of building and staffing a datacenter.
Scalable Across Jurisdictions
The architecture is jurisdiction-agnostic by design. As the Riveon Group expands into Sweden, and evaluates opportunities in the United States, the same model applies: a local operating entity manages the site, Riveon AG provides and finances the hardware, and B2B clients contract for capacity. Each market gets its own operator, its own regulatory footing, and its own service agreements — while the underlying economic logic remains consistent.
GGL Finland holds no custody of any digital assets produced by compute workloads. It does not trade, hold, or settle cryptocurrency. Its revenue is derived entirely from the provision of managed compute services, invoiced and settled in fiat currency.
Infrastructure as a Service Layer
In an industry where vertical integration is often presented as a virtue, the Riveon Group has taken the opposite view. Disaggregating ownership, operations, and end-use creates cleaner audit trails, more defensible tax positions, and a model that institutional investors and regulators can evaluate on its own terms. The CLU-based service fee is legible to any CFO. The operating structure is legible to any regulator.
That is the point. Infrastructure should be boring in its structure and extraordinary in its execution. The Managed Compute Services model is designed to be both.