Microsoft credits custom AI chips with 40% efficiency boost in Azure

Microsoft’s fourth‑quarter fiscal 2026 results highlighted a sharp rise in Azure revenue, which grew 43% year over year, according to the company’s earnings release on July 29. CEO Satya Nadella said the surge is partly driven by the firm’s own artificial‑intelligence processors, which deliver roughly 40% more performance per watt than competing hardware. The higher energy efficiency translates into lower operating costs for Azure and could improve the cloud segment’s profit margins.

Historically, Microsoft has depended on external suppliers such as Nvidia for the compute power needed to run large language models, including those from OpenAI and Microsoft’s internally built MAI series. In recent years the tech giant has shifted toward its proprietary chips, a move Nadella described as “doubling down on custom chips.” The company claims that this strategy is already yielding tangible cost savings and is expected to sustain revenue and earnings growth as demand for AI‑driven cloud services continues.

The financial filing also revealed a $678 billion cloud backlog at the end of the fiscal year, an 84% increase from the prior year. Analysts project AI infrastructure spending could approach $1 trillion within three years, positioning Microsoft to benefit from the expanding market while its custom silicon improves Azure’s economics. Despite the strong performance, Microsoft’s stock has underperformed the broader S&P 500, which rose 12% versus Microsoft’s 5% gain to date.

For investors, the efficiency gains from Microsoft’s in‑house AI chips suggest a competitive edge in the rapidly growing cloud‑AI sector, potentially leading to higher margins and stronger long‑term growth for the company’s cloud business.