Advancing Open Modular Liquid Cooling for Dense Compute at Scale

Explore Supermicro's second-generation direct-to-chip liquid cooling architecture (DLC-2) designed for high-density AI accelerator deployments in this 16-minute conference talk. Learn how next-generation liquid cooling implementations address the critical challenge of efficiently cooling dense GPU clusters within Open Rack environments while maximizing energy, space, and water efficiency. Discover practical insights from real-world facility integrations across diverse system architectures and understand how this technology aligns with the Open Compute Project's objectives for open collaboration, thermal interoperability, and sustainable infrastructure development. Gain valuable knowledge about scaling liquid cooling solutions to meet the demanding thermal requirements of modern AI workloads and dense compute environments.