Professional
background
Head of the State Data Center complex at Kragujevac, Serbia. Electrical engineer with more than 23 years of experience in mission-critical infrastructure, engineering, and operations, played a key role in the development, commissioning, and operation of Serbia’s State Data Center complex.
His expertise includes data center operations, critical infrastructure, resiliency, business continuity, cloud infrastructure, and large-scale digital transformation projects. Led and contributed to major initiatives including the deployment of a 14MW data center campus, EN 50600 Class 4 certification, Oracle Cloud Infrastructure integration, and the implementation of national AI and HPC platforms.
Actively involved in projects related to advanced cooling technologies, energy efficiency, and data center waste heat recovery.
Speaking
session details
Upgrading Data Center for the AI Era: Lessons from Deploying Liquid-Cooled HPC Infrastructure
As AI workloads continue to drive unprecedented power densities, data center operators face new challenges in infrastructure design and operations. This presentation will share practical experiences from upgrading a live EN 50600 Class 4 (Tier IV-equivalent) data center environment to support next-generation AI and HPC platforms based on liquid-cooling technologies.
The session will cover key aspects of planning and implementation, including power and cooling readiness, risk management, operational continuity during deployment, redundancy considerations, and the decision-making process behind balancing design principles with the specific requirements of high-density AI infrastructure.
Attendees will gain insights into real-world challenges, lessons learned, and best practices for integrating liquid-cooled AI systems into mission-critical environments while maintaining high levels of availability, resilience, and operational efficiency. The presentation will also explore the growing impact of AI on future data center design, cooling strategies, and sustainability initiatives.