The power behind Europe's AI.
We run NVIDIA Vera Rubin clusters on our own renewable energy — the lowest cost per token, on European soil, with full data residency.
What scale can't replicate
Hyperscalers and large neoclouds win on scale. We win on structure — four advantages no one closes by buying more GPUs.
We own the power
Dedicated solar and storage sit behind every cluster. Near-zero marginal energy cost and a fixed price that doesn't ride the electricity market — because we don't buy from it.
Time-to-power, solved
Land, grid connection and generation are ready in Extremadura while saturated hubs face multi-year queues. The bottleneck isn't chips — it's connected power, and we have it.
Sovereign by design
Spanish jurisdiction, EU data residency, ENS security certification track. Workloads in health, defence and regulated finance that US-domiciled clouds simply cannot serve.
Direct, at your scale
Founder-level relationships and response times. We treat a growing AI company as a partner, not as ticket #4000 in a hyperscaler support queue.
Three ways to consume our capacity
Pure infrastructure — model-agnostic, no first-party applications. You keep full control of your stack, models and data.
Reserved capacity
Multi-year, take-or-pay blocks of dedicated GPUs. Predictable, bankable capacity for teams that need guaranteed access at a fixed price.
GPU-as-a-Service
On-demand and short-term Rubin GPU-hours with EU data residency. Flexible capacity for inference and agent workloads, without building your own infrastructure.
Sovereign dedicated pods
Physically isolated, ENS-certified capacity for health, defence, finance and government workloads that require sovereignty and auditability.
Firm pilot, committed scale-up
We separate what is operating now from what is committed and phased — and we deploy capacity at the pace demand and financing justify.
Across Extremadura and Andalucía, backed by two regional governments.
Institutionally anchored
Built with partners who give the project its foundation — from silicon to regional government.
Reserve European Rubin capacity
Talking to AI and agent companies, integrators and public-sector teams who need sovereign EU inference at the lowest cost per token. Let's discuss capacity for Q4 2026 and beyond.
Talk to us