Verda Raises $189M Series B at $165M ARR, Targets 250MW of AI Compute by 2027
Verda closed a $189 million Series B on Tuesday, bringing total financing past $450 million and one of the largest funding rounds ever closed by a European independent AI compute provider.
The round was oversubscribed. Emergence Capital led; MUFG Innovation Partners, Supermicro, Varma Mutual Pension Insurance Company, Lifeline Ventures, 6 Degrees Capital, byFounders, and Tesi (Finnish Industry Investment Ltd) participated alongside angels including Ola Tørudbakken and ML engineer Mark Saroufim.
The revenue case
Verda reported $165 million in annualised revenue as of July 2026 — a number that justifies the capital raise on growth-stage terms rather than pre-revenue ambition. The company operates across 50-plus countries with more than 250 staff in Helsinki, London, Taipei, and San Francisco.
CEO Ruben Bryon offered a blunt frame for the timing: “There’s a window right now to build one of the defining compute companies of this generation, and to do so from Europe. It won’t be open for long.”
What the capital unlocks
Verda builds its own data centres, the infrastructure inside them, and the software platform on top — the same vertical integration thesis CoreWeave took public. The new capital deepens all three.
On the hardware side, the company expects to operate more than 250 MW of live capacity by 2027, with Finland active today and additional builds across Europe, the UK, the US, and Asia. It anticipates early deployments of NVIDIA VR200 NVL72 — the NVLink rack system for the next generation of GPU clusters — in the coming months.
On the software side, a container registry has already shipped. Instant clusters already ship with Kubernetes pre-configured for Kueue or Slurm via Slinky. S3-compatible object storage and a dedicated Managed Kubernetes offering are the next items on the roadmap.
Customer mix
Current anchor customers span research and commercial AI: Aleph Alpha runs its R&D infrastructure on Verda, with engineers-in-residence working on the underlying software stack. Magnific handles media generation at millions of requests per day from Verda clusters. Epsilon Health trains radiology models at native image resolution on a dedicated build.
The AI Lab initiative funds co-research with model labs and open-source projects, explicitly framed as a way to understand what compute is actually being used for — and feed that back into data centre design and hardware selection.
The European compute gap
Verda’s fundraise lands as European AI infrastructure policy has moved from aspiration to capital deployment. The EU AI Act is in force; member states are actively competing for hyperscale-scale anchor tenants. Verda’s Finnish base provides natural advantages in power access and cooling costs that US-based rivals cannot replicate from domestic greenfield builds.
The Supermicro participation is worth noting. Supermicro manufactures GPU server hardware and has faced its own turbulence over the past two years; joining a European compute raise signals strategic interest in distribution channels outside the standard hyperscaler order flow.