Big Blue Poised to Peddle Lots of on Premises GenAI (www.nextplatform.com)

IBM is positioning itself to lead in the on-premises generative AI (GenAI) market by enhancing its hardware and software offerings tailored for enterprise use. CEO Arvind Krishna emphasized that while enterprises are excited about GenAI, they are proceeding with caution, unlike the Dot Com Boom's reckless enthusiasm. IBM has spent nearly a decade developing tensor math engines for its Power and Z-series compute platforms, which are critical for handling mission-sensitive data. With its Power 11 and Z 17 systems featuring improved tensor cores, IBM is poised to meet increasing demand as enterprises consider deploying 50% of their AI capabilities in private clouds or data centers. Significantly, Krishna forecasted that as much as 90% of AI inference might occur on-premises using smaller models, potentially leading to cost efficiencies despite higher public cloud spend. IBM's focus on integrating its own models into tools like WatsonX and providing code assistance to modernize legacy applications demonstrates its commitment to becoming a pivotal player in GenAI adoption within enterprises. The company reported cumulative AI-related consulting and software bookings of over $12.5 billion since Q3 2023, highlighting a solid, if still evolving, revenue stream within its broader business context of $100.8 billion in revenues.