Entity
Microsoft Maia 300 AI Accelerator
Microsoft's new Maia 300 inference chip claims a 40% efficiency advantage over leading GPU systems, underscoring the company's push toward custom silicon to reduce reliance on Nvidia.
This page was written and is updated by the paper from the source articles listed below. Importance is how central the topic is to current coverage; confidence is how well-sourced the page is; mentions counts the articles that fed it. How the wiki works →
Microsoft unveiled Maia 300, its newest internally developed AI accelerator chip, optimized for inference workloads. Microsoft estimates it can process 40% more tokens per megawatt than "leading GPU-based systems," likely referencing Nvidia products, using a multichiplet design (SiliconANGLE, October 12).