Delta at OCP Global Summit 2026

Speaking Sessions

Join our industry experts for insights into the latest technologies, design strategies, and industry and market trends shaping the next generation of scalable, efficient, and sustainable AI data centers.

Executive SessionOct 12, 1:40 PMLL20BCD

Delta 800 VDC Power and Cooling Solutions for Data Centers

Join this session to learn how Delta's 800 VDC architectures, Cap-Shelf power buffering, and liquid cooling solutions enable you to solve today's rapidly growing AI data center challenges through greater efficiency, reliability, and scalability. Attendees will leave with a clear view of how power and thermal design are co-evolving to underpin the next generation of AI infrastructure.

800 VDC distribution is increasingly the modern AI data center's best option. This session shows how Delta's 800 VDC and ±400 VDC architectures cut conversion stages between grid and chip, reducing energy losses and freeing up rack space for compute. Cap-Shelf power buffering manages the transient loads created by dense GPU clusters, smoothing demand spikes that strain conventional UPS designs. Paired with liquid cooling built for high-density racks, these architectures enable operators to push power density higher without sacrificing reliability.

Ralf Pieper, Research and Design Director - Custom Design Business Unit

Chih-Chung Chen, Sales Product Manager - Liquid Cooling Solution Business Unit

Expo Hall SessionOct 13, 12:50 PMExpo Hall Theater

From Grid to Chip: Building the Next Generation AI Factory

In this session, we'll explore how integrated power, cooling, and control architectures enable AI data centers to handle rising power densities while enabling more efficient, scalable, and resilient next-generation AI compute environments.

Why are operators treating power, cooling, and control as one integrated system rather than three separate disciplines? Join us to see how coordinated architecture decisions at each stage compound into measurable efficiency and uptime gains from grid to chip. Attendees will leave with a clear understanding of how such unified systems enable facilities to manage rising power densities without the efficiency losses and reliability risks that come from designing each layer in isolation. The result is an infrastructure blueprint built for the density, speed, and resilience the next generation of AI compute now demands.