Europe's Nordic Unicorn: Verda Raises $189M to Build a Full-Stack AI Cloud
Helsinki-based Verda raised an oversubscribed $189M Series B led by Emergence Capital, valuing the full-stack AI cloud at over $1B — with Supermicro on the cap table, 250 MW planned for 2027, and a large-scale inference service next on the roadmap.
Helsinki, 10am EEST, Tuesday September 22 — European technology company Verda announced $189 million (€163 million) in new funding, an oversubscribed Series B led by Emergence Capital that pushes the company’s valuation past $1 billion and makes it Europe’s — and Finland’s — latest AI infrastructure unicorn. For a continent still searching for an answer to America’s hyperscalers, the round is being read as one of the clearest signals yet that Europe intends to compete on compute rather than concede it.
The financing brings Verda’s total raised to more than $450 million across equity and debt. Beyond lead investor Emergence Capital — the Silicon Valley firm known as an early backer of Salesforce and Zoom — the round pulled in an unusually strategic mix: MUFG Innovation Partners, server-maker Supermicro, Varma Mutual Pension Insurance Company, Lifeline Ventures, ENDUR, 6 Degrees Capital (6DC), byFounders, Tesi (the Finnish Industry Investment fund), and angels including Ola Tørudbakken and Mark Saroufim of Core Automation and GPU MODE.
What Verda actually builds
Founded in Helsinki in 2020 by Ruben Bryon, Verda (formerly DataCrunch) is a full-stack AI cloud company — a phrase it uses precisely. The company controls every layer of the stack: its own physical data centers, the compute hardware inside them, the cloud platform software, and even an in-house AI Lab that contributes to frontier research and open-source projects. On the product side today that means fast-access GPU instances on NVIDIA’s latest silicon (GB300 NVL72, B300 SXM6, B200 SXM6, RTX PRO 6000, H200, H100, A100), self-service InfiniBand-connected GPU clusters, auto-scaling serverless GPU containers for inference and batch jobs, NVMe block storage, and hardware-attested confidential computing for inference and fine-tuning.
That breadth is the bet. Verda’s pitch is that in the agentic era, frontier AI teams do not want to stitch together a data-center lease, a hardware vendor, a scheduling layer, and a serving stack — they want one provider that owns the whole chain and can move capacity fast. “Demand for AI compute is insatiable, and the gap between demand and supply is growing,” said Joe Floyd, General Partner at Emergence Capital. “Verda is moving quickly to secure and scale capacity, with the technology, ambition, and momentum to become a defining AI infrastructure company.”
The numbers behind the round
The commercial trajectory explains the oversubscription. Verda says it reached a $165 million annualized revenue run rate in July 2026 — up from roughly $100 million just two months earlier — and now serves organizations in more than 50 countries. The company employs around 250 people drawn from over 40 nationalities, and recently opened offices in London and San Francisco with further expansion planned across Europe, the UK, the US, and Asia.
The fresh capital is earmarked for three moves. First, launching a large-scale inference service, aimed at the workload layer where the agentic era’s economics are increasingly decided. Second, multiplying compute capacity over the next year: Verda currently operates data-center capacity in Finland and expects early deployments of NVIDIA’s VR200 NVL72 systems in the coming months, with more than 250 MW of operations planned for 2027. Third, deepening the platform — S3-compatible object storage, managed Kubernetes, enterprise features, and continued investment in performance engineering and the AI Lab.
Why Europe is watching closely
The subtext of the announcement is geopolitical as much as commercial. European policymakers and enterprises have spent two years worrying aloud that the continent’s AI ambitions are hostage to US cloud capacity — a dependency felt most acutely by banks, governments, and manufacturers with data-residency obligations, and sharpened by the energy and sustainability constraints that make Nordic locations attractive. Verda runs its data centers on 100% renewable energy, and its founder frames the mission in civilizational terms.
“AI is becoming critical infrastructure across industries, and the next few years are a pivotal window for Europe,” said Ruben Bryon, Verda’s founder and CEO. “Our north star is to build the first true tech company in Europe, with AI infrastructure as the starting foundation, and to bring down the carbon footprint of compute worldwide. We’re proud of our progress but are a long way from being done.”
That ambition — “the first true tech company in Europe” — is deliberately provocative. Europe has produced global companies, but rarely ones that own a foundational technology layer end-to-end at scale. Whether a Helsinki operator can hold that position against AWS, Azure, Google Cloud, and a rising generation of US GPU clouds is an open question. Verda’s answer is specialization and speed: full-stack control, Nordic power economics, EU data residency, and the flexibility to deploy next-generation NVIDIA infrastructure months rather than years after announcement.
The competitive landscape
Verda is not alone in the European GPU-cloud race — Scaleway, OVHcloud, Nscale, Hetzner, IONOS, and Exoscale all compete for some slice of the same workloads, and each has responded to the AI boom with its own GPU offerings. Verda’s differentiators are the depth of its stack, its early access to cutting-edge hardware (Supermicro’s presence in the round is a supplier relationship as much as an investment), and its cash-flow-positive operating history — a rarity among infrastructure startups that typically burn capital for years before reaching Verda’s revenue scale.
The next test arrives with inference. Training capacity grabs headlines, but 2026’s fastest-growing demand is serving — agents and applications that need GPU compute continuously and cheaply, with latency constraints that make distributed European capacity genuinely valuable. Verda’s planned large-scale inference service, backed by 250 MW of planned capacity, is positioned to be the centerpiece of that effort.
For now, Europe has a new unicorn with a balance sheet to match its ambition — and a founder openly stating that the goal is nothing less than rebuilding the continent’s standing in technology from the infrastructure up.
Sources
- [1] https://verda.com/blog/verda-raises-189m
- [2] https://pulse2.com/verda-raises-189-million-series-b-to-become-europes-latest-ai-infrastructure-unicorn/
- [3] https://www.vestbee.com/insights/articles/verda-raises-189-m
- [4] https://dealroom.co/news/155335-verda-hits-160m-series-b-becomes-finlands-newest-unicorn/