Microsoft Deploys Maia 300 as In-House Nvidia Alternative, Shifting AI Chip Strategy to Custom Silicon

Microsoft announced Maia 300, its third-generation AI accelerator, on October 12, 2026, marking an escalation in the company's effort to reduce dependence on Nvidia's GPUs by building proprietary silicon at scale. The chip will begin deployment in 2027 and become the primary inference system for Microsoft's products and services, CEO Satya Nadella stated, operating at multi-gigawatt power levels across Azure data centers.