Maia 200 is already serving Microsoft Copilot and the MAI team; Microsoft says GitHub Copilot and Microsoft Foundry are to follow by year-end.
Maia 300 is scheduled to begin production deployments in 2027. Microsoft describes it as an inference-first accelerator for long-context, memory-intensive and decode-heavy workloads. Each chip has eight HBM4 stacks, 288 GB of memory, 18 TB/s of bandwidth and a 950-watt TDP. A 96-accelerator rack is specified at 167 kW, with 27.6 TB of HBM and 1.7 PB/s aggregate bandwidth.
Microsoft’s headline expectation is 40% more tokens per megawatt than leading GPU systems. InfoQ notes that the company has not identified the comparison GPU, workload, context length, latency target, or definition of a useful token. The report also says Microsoft has not stated whether Azure customers will be able to provision Maia directly. The source does not establish independently measured performance, cost, or customer availability.