Microsoft Plans Increased Use of Self-Developed AI Data Center Chips

Microsoft is shifting its data center strategy to rely primarily on its own custom-designed chips to support AI workloads, aiming to reduce dependence on major chipmakers like Nvidia and AMD. The company has already introduced in-house chips such as the Azure Maia AI Accelerator and the Cobalt CPU and is developing next-generation versions alongside innovative cooling technologies like microfluids. Chief Technology Officer Kevin Scott highlighted that this move allows Microsoft to optimize performance by controlling the entire system architecture, including chips, networks, and cooling. Despite significant investments and rapid capacity expansion, demand for AI computing continues to outpace supply, pushing Microsoft and other tech giants to increase spending on AI infrastructure. Microsoft's approach reflects broader industry trends as companies seek to fine-tune AI performance and manage the growing computational demands more efficiently. Analysts maintain a strong buy consensus on Microsoft stock, reflecting confidence in the company's strategic direction and financial strength.