Developing story · 2 episodes · 12 Oct to 17 Oct
Microsoft's Custom Chip Push for AI Efficiency
Microsoft is advancing its custom silicon strategy by unveiling the Azure Maia 300 chip, which promises significantly higher efficiency for AI workloads. The company is now considering making this new chip its primary inference engine due to its superior memory bandwidth capabilities.
-
Microsoft could make Maia 300 its primary inference engine with 2.5x memory bandwidth
Microsoft announced that its upcoming Maia 300 accelerator will become the main inference engine for Azure, targeting multi‑gigawatt deployments beginning in 2027. The chip swaps a modest 36%…
1 source -
Microsoft unveils Azure Maia 300 chip for 40% more tokens per megawatt in 2027 deployments
{'Microsoft announced its third-generation Azure Maia 300 AI accelerator, set for production in 2027, targeting multi-gigawatt deployments. The chip aims to deliver 40% more tokens per megawatt than…
12 sources primary source HN 7
Who and what
MicrosoftOpenAITSMCGitHubNvidiaGoogleAmazonGartnerAzure Maia 200Azure Maia 300MAI modelsScott Guthrie