Ai predicts data center heat to improve cooling and design
AI-driven Thermal-aware Data Center Capacity Planning
Machine Learning
Summary
Data centers that run big AI models use a lot of power and create hot spots, which are tricky to cool efficiently. The authors created an AI system that quickly predicts how hot different parts of a data center will get based on many factors like server layout and cooling settings. This new system matches the accuracy of slow, detailed simulations but runs thousands of times faster. Because it works so quickly and accurately, data center managers can instantly find better ways to distribute work and manage cooling.
What this means in practice
- •For data center operators: Optimize workload distribution and HVAC settings instantly to prevent hotspots and enhance cooling efficiency in real data center designs.
- •For it infrastructure planners: Plan data center capacity quickly and accurately by predicting thermal effects of various server layouts without costly detailed simulations.
Authors
Yixing Li, Mark Fenton, Matthew Kaufeler, Ka Ming Leung, Xin Ai, Zhiyu Zeng
Abstract
The emerging of large language models (LLMs) has posed significant challenges to the thermal management of data center. Intense GPU computation for LLMs results in localized hotspots. Moreover, spiking thermal loads during training and inference bursts make real-time cooling response more difficult to predict and control. Thermal-aware capacity planning of data center requires massive expensive high-fidelity CFD simulations. AI models can perform real-time prediction for unseen designs. However, existing works either have large prediction error, or have over-simplified assumptions for data center operations. This work presents an AI-driven framework that can perform thermal-aware capacity planning for a real-world data center in seconds. The embedded AI model learns from numerous key parameters (rack power, server power, server placement, HVAC settings etc.), and provides temperature prediction within milliseconds. This AI model is tested against high-fidelity CFD simulations, and results show that for unseen data center designs, model can achieve high accuracy with 10000X speedup. Driven by the AI model, the authors design the thermal-aware capacity planning framework. This framework can help data center designers and operators instantaneously optimize both workload distribution and HVAC cooling efficiency.