OVHcloud flags revenue hurdle for EU AI gigafactories
Europe's plan to build up to seven AI gigafactories faces a commercial viability challenge as operators weigh the heavy cost of graphics chips and limited public demand. OVHcloud Chief Executive Octave Klaba says annual sales of 300 million euros to 400 million euros are needed just to break even, far above the level implied by current public commitments.
Highlights
- OVHcloud chair Octave Klaba warns that the proposed 200 million euro public procurement over five years is insufficient to support AI gigafactory investment, which requires at least 600 million euros for chips alone.
- Europe's planned AI gigafactories need 300 million euros to 400 million euros in annual revenue to break even, but public sector orders would only cover 10% to 15% of this requirement.
- Klaba states that European-scale operations are essential, as national-level demand is too weak to make gigafactory investments profitable and attract sufficient private capital.
Funding gap clouds deployment plans
As reported by Reuters, Klaba says a proposed 200 million euro public commitment to buy AI services over five years is too small to support the scale of private investment needed for the facilities. He says the public sector would buy about 40 million euros of computing services a year, while an operator would spend at least 600 million euros on chips alone.Klaba says Europe's proposed gigafactories, designed for training, fine-tuning and running AI models, would need 300 million euros to 400 million euros in annual revenue to break even. That means the public order would cover only about 10% to 15% of the turnover required, as the underlying graphics hardware depreciates quickly.
European scale seen as essential
The EuroHPC program is seeking industry consortia to build and operate as many as seven AI gigafactories across Europe. Klaba says OVHcloud has not joined existing consortia because a focus on national markets limits the commercial case for projects of this size.He argues that national demand is too weak to make such investments profitable and says the projects need to operate on a European scale to attract enough customers. His comments point to a broader challenge for the region's AI infrastructure push, where public policy ambitions may need larger cross-border demand support to unlock private capital.
In our earlier article on the surge in oversized Series A rounds, we highlighted how early-stage venture investors are writing much larger checks, with AI-focused startups taking the majority of $100 million-plus deals in 2026. We also noted that this funding concentration reflects sustained confidence in scaling AI products and infrastructure, even as the market remains selective about which teams and business models it backs.