#Azure
- Microsoft's Maia 300 Bets on Memory, Not Muscle, to Cut AI Costs
Microsoft will make its own Maia 300 chip the primary engine for AI inference at multi-gigawatt scale in 2027. The bet is on memory bandwidth, not raw compute.
Microsoft will make its own Maia 300 chip the primary engine for AI inference at multi-gigawatt scale in 2027. The bet is on memory bandwidth, not raw compute.