Reproduce “AI Surge Accelerates the Transition to Liquid Cooling: Rethinking Energy and Water Use in Data Center Cooling” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yingbo Zhang, Eric Masanet, Shengwei Wang
The authors model single-phase cold-plate cooling across climates, comparing evaporative and dry heat rejection with air cooling. Both liquid options use less cooling energy; evaporative cooling saves 33–56% at settings. Thresholds of 33°C and 49°C achieve at least 99.5% annual free cooling. Warmer coolant can raise GPU power, so total-energy optima vary by climate; these are modeled results, not guarantees.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.