Microsoft is moving AI inference off Nvidia and onto itself

Microsoft will increase its use of in-house AI chips, according to CNBC Tech and Reuters. The company aims to rely less on Nvidia and cut costs by using its own hardware for AI inference.

Why it matters

Switching to homegrown chips could help Microsoft expand its profit margins and control more of its AI supply chain. Microsoft has invested heavily in AI infrastructure and is now moving to internal solutions, as reported by CNBC Tech. For businesses, this shift signals a possible change in the availability and pricing of AI services built on Microsoft's cloud.

Microsoft's new strategy for AI chips

Microsoft will mainly use its Maia chips for AI inference, according to CNBC Tech. The company plans a large-scale deployment, Reuters reports, which is expected to reduce expenses tied to buying chips from Nvidia. Microsoft has not said when the switchover will be completed or how much of its infrastructure will use these in-house chips. The move is part of a broader trend among big tech firms to bring critical AI hardware in-house, as seen also in Amazon and Google efforts. By doing this, Microsoft keeps more control over supply and cost, both of which have been under pressure due to global chip shortages and strong demand for AI workloads.

What this means if you run a business

For company owners and operators, Microsoft's decision could affect the cost and supply of AI-powered services. If you rely on Azure for AI workloads, changes in chip sourcing may impact pricing, availability, or performance down the line. This move could also influence contract terms or service-level agreements with Microsoft, depending on how the transition plays out. Businesses developing their own AI products might see a shift in which hardware is recommended or supported by Microsoft's cloud. If you are considering long-term AI investments or partnerships, it's worth keeping an eye on how quickly Microsoft can scale its chip production and whether other cloud providers follow a similar path.

For more context on Microsoft's broader AI strategy, see our coverage on Microsoft's call for AI regulation.