IBM bets on commodity models for AI edge

IBM bets on commodity models for AI edge
Despite public skepticism and a reported shortfall in OpenAI's annualized revenue, major AI players including IBM, Nvidia, and Marvell are pressing forward. IBM is positioning itself around a four-part architecture of agents, data, automation, and hybrid infrastructure, betting on model commoditization and Red Hat's open-source foundation rather than proprietary models. Analysts at theCUBE Research see IBM's opportunity in owning the operational layer around AI. Meanwhile, Marvell targets $70-90 billion in revenue by fiscal 2031, and the broader market is shifting from AI training toward inference workloads.
Read the original article →