Microsoft (NASDAQ:MSFT) has spent the past two years writing some of the biggest checks in the tech industry as it builds out the infrastructure needed to support the AI revolution.
While investors have closely tracked the company’s soaring capital expenditures, a new deep dive from Morgan Stanley analyst Keith Weiss suggests the market may not fully appreciate what those investments could ultimately be worth.
After analyzing Microsoft’s datacenter expansion and comparing cloud revenue growth to capacity additions, Weiss came away “bullish on upside potential,” arguing that expectations for the revenue those investments can generate may still have room to move higher. The analyst believes Microsoft is building AI capacity ahead of demand, creating the potential for cloud revenue forecasts to rise as monetization catches up.
Using a “revenue per megawatt” framework to evaluate datacenter productivity, Weiss estimates Microsoft’s cloud ecosystem currently generates roughly $20 million to $30 million in annualized revenue per megawatt of installed datacenter capacity. That figure is expected to trend lower over the next several years, reaching the high teens by fiscal 2028.
At first glance, declining revenue per megawatt might look concerning. However, Weiss does not actually view that trend negatively. “We do not interpret this dynamic as evidence of a deteriorating monetization opportunity, but rather as an indication that Microsoft is deploying AI-specific capacity ahead of the associated monetization cycle,” he explained.
The scale of Microsoft’s expansion is significant. Based on the estimates, the company’s installed datacenter footprint could grow from roughly 5 gigawatts in fiscal 2024 to around 20 gigawatts by fiscal 2028. Microsoft is also expected to exit calendar 2025 with approximately 7-9 gigawatts of installed capacity.
Weiss’s analysis goes beyond Azure alone and incorporates Microsoft’s wider cloud ecosystem, including Microsoft 365 Commercial Cloud, Dynamics 365, LinkedIn’s commercial operations, and other cloud services. The reasoning is that Microsoft’s infrastructure supports a “unified cloud and AI platform across Azure and multiple 1st party applications (rather than just Azure/Azure AI).”
Leave a comment