Microsoft

Technology

• Microsoft researchers' Fork-and-Flush method lifts agent results up to 66%: Microsoft researchers found autonomous AI agents get trapped in "idea basins" as chat context grows, local solution regions they cannot escape. Their Fork-and-Flush method clears chat history while keeping the workspace, achieving up to 66% better results on the same compute budget.

  AI Research • The Decoder

  Permalink

• Microsoft plans to scale in-house Maia chips to multi-gigawatt AI inference by 2027: Microsoft plans to make its in-house Azure Maia chips the main AI inference system starting in 2027, at multi-gigawatt scale. The upcoming Maia 300 is expected to deliver 40% more tokens per megawatt than leading GPU systems. CEO Nadella called it Microsoft's biggest bet on its own silicon.

  AI Infrastructure • The Decoder

  Permalink