Problem
Research questions and friction points this paper is trying to address.
thermal management
data center capacity planning
large language models
hotspot prediction
CFD simulation
AI-driven Thermal-aware Data Center Capacity Planning
🤖 AI Summary
This study addresses the challenge of localized thermal hotspots and abrupt load fluctuations induced by large language model (LLM) computations, for which conventional computational fluid dynamics (CFD) simulations are prohibitively expensive to support real-time responses. We propose an embedded AI-based thermal-aware capacity planning framework for data centers that performs multi-parameter feature learning by fusing rack power consumption, HVAC configurations, and other operational variables, thereby overcoming the high error rates and oversimplified assumptions of existing models. Validated against high-fidelity CFD benchmarks, the proposed model achieves highly accurate millisecond-level temperature predictions on unseen designs, delivering a 10,000-fold speedup over traditional CFD. This computational efficiency enables second-level workload distribution optimization and instantaneous decision-making for cooling strategies, offering a practical solution for thermally resilient LLM infrastructure management.
📝 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.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Thermal-aware Capacity Planning
AI-driven Framework
Computational Fluid Dynamics (CFD)
Data Center Cooling
Large Language Models (LLMs)
🔎 Similar Papers
No similar papers found.
