Microsoft Announces New AI Chip Azure Maia 300
Microsoft said Maia 300 will deliver 40% more tokens per megawatt than leading GPU systems as it shifts toward in-house inference hardware.
- On Monday, Microsoft unveiled the Maia 300, an AI chip designed to serve as the primary AI inference system for the company's products and services.
- Seeking improved performance, Microsoft designed the architecture to handle emerging workloads including long-context reasoning, latency-sensitive conversational experiences, and multi-step agentic tasks.
- Built on TSMC's N2P process, a full rack of 96 Maia accelerators delivers up to 7 times the rack-scale token throughput of the Maia 200 model.
- Executive Vice President Scott Guthrie said the chip will "deliver 40 percent more tokens per megawatt" compared to leading GPU-based systems, with production starting in 2027.
- Evercore ISI analysts estimate Maia will represent about 10% of Microsoft's chip spending by early 2027, though the company maintains partnerships with Nvidia and AMD for training models.
62 Articles
62 Articles
Microsoft moves to reduce Nvidia dependence with new Maia 300 AI chips, shifts more AI workloads in-house - Tech Startups
Microsoft wants to run more of its artificial intelligence business on chips it builds itself. The software giant on Monday unveiled its third-generation Maia 300 AI accelerator and announced plans to shift much of the computing behind its AI products […] The post Microsoft moves to reduce Nvidia dependence with new Maia 300 AI chips, shifts more AI workloads in-house first appeared on Tech Startups.
Microsoft Bets Billions on Maia 300 to Power Its AI Future and Curb Nvidia Dependence
Microsoft unveiled Maia 300, promising 40% more tokens per megawatt and 7x rack throughput versus its predecessor. The custom AI chip targets inference for Copilot and agents with massive HBM4 memory. Deployments scale to multi-gigawatts in 2027, reducing reliance on Nvidia GPUs.
New accelerator expected to deliver 40% more tokens per megawatt than leading GPU systems
They seek to reduce their dependence on third-party processors from companies such as Nvidia and AMD
Coverage Details
Bias Distribution
- 48% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
