Neoclouds vs Hyperscalers: Who is Building the Future of European Cloud?

OVHcloud just raised 400 million euros. Scaleway announces 3 new regions. STACKIT, Lidl's cloud subsidiary, invests 1 billion in Germany. European neoclouds are on the offensive. But against AWS (32% of global market), Azure (23%) and Google Cloud (10%), can they really compete? The answer is in the datacenters.

The European cloud market is worth 127 billion euros in 2026. American hyperscalers hold 65%. European neoclouds hold 12%. The rest is shared between Chinese clouds (15%) and national players (8%). These figures hide a more complex reality: neocloud growth is 35% per year, compared to 18% for hyperscalers. The catch-up is underway.

But the catch-up is not technological — it's infrastructural. A European neocloud can develop software as performant as AWS. It cannot, however, build datacenters as fast. An American hyperscaler takes 18 months to build a 40 MW region. A European neocloud takes 3 to 4 years. The difference is not in the code, it's in the concrete, copper and building permits.

Neoclouds: The Shadow Army of Digital Europe

OVHcloud, the largest European neocloud, has 37 datacenters in 19 countries. Scaleway has 8, but with very high density in France. STACKIT, the smallest, has 3 in Germany. IONOS, Hetzner, Exoscale — actors are multiplying. Their common strength: they are European, sovereign, and benefit from growing regulatory support.

GDPR, Data Act, NIS2 directive — each European regulation strengthens the position of neoclouds. A company processing health data cannot legally host it on AWS US-East. It must choose a European cloud. This is the captive market that makes neoclouds' fortune. But this captive market is limited: it represents only 20% of the total cloud market.

Hyperscalers: Defending the Empire

AWS, Azure and Google are not passive. They invest 15 billion euros per year in Europe. They open regions in Strasbourg, Madrid, Milan. They sign partnerships with sovereign clouds (Telefonica with AWS, Deutsche Telekom with Azure). They create 'trusted clouds' — legally European entities that host data on European soil, but remain controlled by American shareholders.

This strategy is formidable. It allows hyperscalers to remain competitive in the sovereign segment without renouncing their business model. A 'trusted cloud' AWS charges 20% more than standard AWS. But it is 30% cheaper than equivalent OVHcloud. For a company looking to reduce costs while remaining compliant, the choice is obvious.

The Investment Thesis: Datacenter as Arbiter

The battle between neoclouds and hyperscalers will not be won in software — it will be won in datacenters. Whoever controls the physical infrastructure controls costs, latency, sovereignty. And this is where European neoclouds have a structural advantage: they are closer to users, more flexible on configurations, and better integrated with local electrical networks.

FINXIA Capital has identified a specific opportunity in this segment: edge datacenters — peripheral datacenters that host neocloud workloads in secondary cities. Lyon, Marseille, Toulouse, Barcelona, Milan. These markets are neglected by hyperscalers but saturated by local SME demand. A 5 MW edge datacenter, Tier III certified, can generate returns of 12% to 15% — well above the 6% to 8% of hyperscalers in primary markets.

What Investors Must Understand

The cloud war is not a binary battle. Hyperscalers will not disappear. Neoclouds will not replace them. What is emerging is a market fragmentation: hyperscalers will keep large accounts and non-critical workloads, neoclouds will capture the sovereign and regulated segment, and edge datacenters will take the intermediate market of SMEs and local administrations.

The investor who understands this tripartition can position their capital in each segment. But the most underaddressed — and most profitable — segment is the edge datacenter. Assets are small, entry tickets are affordable, returns are high, and demand is structural. This is where FINXIA Capital focuses its active research.