Maia 200 boosts performance and efficiency for large-scale AI workloads

What matters

• Built on TSMC’s 3nm process, delivering over 10 petaFLOPS in FP4 and 5 petaFLOPS in FP8 precision
• Redesigned memory system with 216GB HBM3e at 7 TB/s for efficient data movement
• Seamless integration with Azure, previewing the Maia SDK for model optimization

Why it matters

Seamless integration with Azure, previewing the Maia SDK for model optimization

Source: https://blogs.microsoft.com/blog/2026/01/26/maia-200-the-ai-accelerator-built-for-inference/

Drafted by the GenAI News review pipeline. Verify details against the source before publishing.