AZURE MAIA 300 ARCHITECTURE DEEP DIVE

Summary from the TLDR DevOps newsletter

Microsoft's third-generation AI accelerator uses TSMC N2P with 288 GB HBM4 at 18 TB/s. Racks hold 96 accelerators at 167 kW, scaling to 12,288-chip superpods, with up to 7x token throughput.