NVIDIA partners expand European cloud capacity for sovereign AI
Original title: NVIDIA Cloud Partners Scale Across Europe — Sovereign, Open and Built for Safe AI Agents
Nscale says DSX MaxLPS ran 37% more GPUs and nearly 50% more inference throughput on GB300 NVL72 within the same power budget.
NVIDIA says cloud partners across Europe, the Nordics, Africa, and Central Asia are adding regional training and inference capacity and controls for agentic systems. Nscale reports power-efficiency tests on GB300 NVL72 and plans Vera Rubin deployments across five European markets; Nebius is adopting Groq 3 LPX with Vera Rubin NVL72, and Mistral is expanding accelerated infrastructure, including Vera Rubin planned for 2027. The post also describes NVIDIA OpenShell as a reference architecture for limiting what agents can access, though the description is cut off.
Unconfirmed: Performance figures are partner or customer results reported in an NVIDIA blog, and the OpenShell section is truncated.
- Mistral Large 4 was trained on 3,800 Grace Blackwell GPUs in Mistral-operated European data centers.
- CodeRabbit’s Nemotron 3.5 Lightning routing agreement rose from 75.8% to 80.7% while estimated inference cost fell 50%.
- Nebius put three NVIDIA retail agent blueprints into early access and is a computer supplier for SPRIND’s Next Frontier Challenge in Germany.
- Fireworks and Era4 plan Blackwell Ultra capacity at Era4’s Chesterfield facility for UK and European enterprises.
- Vendors
- NVIDIA
- Countries
- GermanyUnited Kingdom
From the original article
4 min read at blogs.nvidia.com
Demand from AI-natives is pulling the European AI cloud ecosystem forward faster than ever. Developers need compute closer to their data, open models they can customize and serve at scale, and full-stack infrastructure that is secure by design. NVIDIA Cloud Partners across Europe, the Nordics, Africa, and Central Asia are expanding regional compute capacity for […]
By Stuart Pitts · blogs.nvidia.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.