The transition toward liquid cooling represents a fundamental change in how data centers handle the extreme thermal loads generated by modern AI processors. SUNON’s approach integrates thermal technology across three distinct tiers: chip-level cooling, server-level systems, and rack-level architecture. By standardizing these levels, the company aims to offer operators greater flexibility in managing varying power densities and operating environments while maintaining high system reliability.

Attendees at the OCP Global Summit, scheduled for October 12–15 at the San Jose McEnery Convention Center, can view these technologies firsthand at booth A19. The presentation focuses on streamlining the deployment of high-performance computing clusters by addressing the cooling requirements that currently limit hardware density and efficiency in modern AI environments.