Microsoft Leans on Neoclouds to Ease AI Crunch

Investment & Strategy - Microsoft has invested $33 billion in Neo cloud infrastructure, engaging European players like End Scale and Nevius, as well as local player CoreWeave [1] - Microsoft is leveraging Neo clouds to accelerate AI model development, addressing capacity constraints for its customers, itself, and OpenAI [2][4] - Microsoft is utilizing Neo clouds to build its own AI models, reflecting a strategic move to catch up with OpenAI and other AI leaders [4][5] Financial Implications - Renting cloud capacity allows Microsoft to classify these expenses as operating expenditures (OpEx) rather than capital expenditures (CapEx), impacting financial reporting [6][7] - Shifting expenses to OpEx provides financial flexibility, allowing Microsoft to adjust cloud capacity based on evolving needs without being tied to depreciating assets [7][8] Technology & Infrastructure - Microsoft's internal large language model (LLM), developed under Mustafa Suleyman, utilizes CoreWeave's assets, which are based on Nvidia hardware [5][6] - Neo clouds are dedicated to AI tasks, including training, inference, and storage, distinguishing them from general-purpose cloud infrastructure [2][3]