Compal Showcases NVIDIA AI Factory Infrastructure at OCP Global Summit

Compal Electronics showcased its next-generation AI factory infrastructure at the OCP Global Summit 2026 in San Jose, California. The company presented solutions integrating power delivery, cooling technologies, and accelerated computing to address the growing demands of AI workloads.

At the event, Compal emphasized that as AI data centers increase in density and power requirements reach megawatt levels, simply adding more GPUs is insufficient. The entire infrastructure, including power supply, thermal management, and rack integration, needs redesigning for performance, efficiency, and scalability.

Compal is focusing on NVIDIA's accelerated computing solutions, combining high-density compute platforms with next-generation power and cooling technologies. The company highlighted an 800V DC architecture and megawatt-class power infrastructure designed to meet future AI data center needs. This high-voltage DC approach reduces energy conversion complexity and supports increasingly dense AI environments.

Furthermore, Compal introduced advanced liquid cooling technologies and megawatt-scale cooling capabilities to ensure efficient operation of AI systems despite rising rack density. The company views compute, power, and cooling as interconnected components of the AI factory, optimizing the infrastructure for high-performance accelerated computing.

"Building the next generation of AI-driven factories requires us to rethink the entire infrastructure surrounding accelerated computing," said Alan Chang, Vice President of ISBG at Compal. "Compal is combining power, cooling, rack integration, and NVIDIA's accelerated computing to help customers build AI infrastructure that scales efficiently, from the platform level to megawatt deployments."