SemiAnalysispodcast

AgentX - InferenceXv3: Does CUDA Moat Hold up in Agentic Inferencing?

Monday, August 24, 2026Cam QuiliciView original
AgentX - InferenceXv3: Does CUDA Moat Hold up in Agentic Inferencing?
$3 Million USD dataset open sourced, 1 Mil+ Context Length, Multiturn, Sub Agents 95%+ KVCache HitRate, GB300 NVL72, MI355, B200