Christoph Hagleitner

IBM, Switzerland

System Architecture for AI Factories

Abstract

The widespread adoption of AI models in domains like science, health, social media, etc is transforming the infrastructure landscape. Classic on-prem deployments targeted at AI-training and HPC workloads are not well suited for AI-applications that deploy LLM-based recommendation systems (e.g., domain specific chat bots, code assistants), science applications (e.g., AI-driven material science, weather models combining AI-models and physics-based simulations) and other AI-infused workflows. These applications require us to rethink the control-plane and SW-stack for on-prem systems and also have implications for the processors and AI-accelerators at the core of these systems.

Biography

Christoph Hagleitner leads the “Heterogeneous Cognitive Computing Systems” group at the IBM Research –Zurich Lab (ZRL) in Ruschlikon, Switzerland. The group focuses on heterogeneous computing systems for cloud datacenters and HPC. Applications include security, big-data analytics and cognitive computing. He obtained a diploma degree in Electrical Engineering from ETH, Zurich, Switzerland in 1997 and a Ph.D. degree for a thesis on CMOS-integrated Microsensors from ETH, Zurich, Switzerland in 2002. In 2003 he joined IBM Research to work on the system architecture of a novel probe-storage device (“millipede”-project). In 2008, he started to build up a new research group in the area of accelerator technologies. The team initially focused on on-chip accelerator cores and gradually expanded its research to heterogeneous systems and their applications.