Databricks hits $188B valuation as AI infrastructure play

NVIDIA released Nemotron 3.5 Super VL, a 120-billion-parameter open model designed for enterprise agentic workloads, alongside two edge-optimized variants, Bolt 74B-A9B 24GB and Bolt E4B 4GB. Super VL claims 5x faster throughput than comparable models while delivering accuracy comparable to NVIDIA's largest reasoning models at one-fifth the parameter count. The three models span deployment tiers from data centers to local devices, allowing organizations to run agents on their own infrastructure without trading capability for deployability.

Microsoft and Nvidia launched RTX Spark laptops featuring Arm-based chips instead of traditional x64 architecture from AMD and Intel. The new platform combines Nvidia's GPU technology with Arm processors designed for AI workloads. Microsoft has undertaken significant optimization work to prepare Windows for these machines, marking a shift in how the operating system handles non-traditional processor architectures.

Portland-based Mueon raised $65 million to develop a new server rack architecture for AI data centers. The company claims its design can fit up to ten times more computing capacity into the same power footprint as existing setups. The funding targets a critical constraint in AI infrastructure as demand for data center capacity continues to grow.

The Nordic region is establishing itself as a major European AI infrastructure hub, with countries deploying approximately 1 gigawatt of AI cloud capacity by end of 2027 and leveraging advantages in renewable energy and data governance. Major companies including Spotify, ABB Robotics, and Elisa are building AI applications on NVIDIA infrastructure, while startups raised $2.8 billion in the first half of 2026 and are scaling on open models. The region's transition from producing high-density unicorn startups to building foundational AI infrastructure signals a shift from ambition to operational deployment.