Europe's automation leaders are reshaping their business models as demand for AI infrastructure grows. Schneider Electric and Siemens are investing heavily in industrial software, while ABB is still focusing on physical equipment.
NVIDIA has published its entire Run:ai GPU orchestration platform as open source under the Apache 2.0 license. The release includes the KAI Scheduler, GPU fractioning with CUDA 13.4, quota management and a local development environment based on Kind.
Apple has published research on Minions, a system that pools unused GPU resources across clusters of Apple Silicon Macs to run AI inference. The idea is not new, but the work builds on existing approaches and further positions the Mac as an AI machine.
Meta has announced Petal, a subsea cable that will connect the United States to France from 2029. Its capacity is meant to be double that of today's best systems, to carry the growing data traffic driven by AI.
This white paper describes how edge AI and physical AI could let cities go beyond passive monitoring and act on real-time conditions in public safety, mobility and infrastructure. It argues that processing at the edge cuts latency, keeps sensitive data local and reduces reliance on central cloud systems.
After stock crashes and Firmus's cancelled IPO, industry executives expect only seven to ten of the hundreds of AI cloud providers to survive. CoreWeave has lost a quarter of its market value and Microsoft dropped a deal with Nscale for 23,000 Nvidia chips, a risk for Nvidia since customers outside the hyperscalers make up 45% of its data center revenue.