Fusion Intelligence for Digital Twinning AI Data Centers: A Synergistic GenAI-PhyAI Approach

2025-05-26

AI Summary

To address the dual bottlenecks of hallucination in generative AI (GenAI) and poor generalizability in physics-informed AI (PhyAI) for constructing digital twins of AI-dedicated data centers (AIDCs), this work proposes the first GenAI–PhyAI dual-agent collaborative framework. It leverages large language models (LLMs) for natural-language-driven modeling, tightly coupling symbolic physics modeling, multi-scale thermo-fluid dynamic constraints, and real-time data assimilation to enable end-to-end, physics-consistent digital twin generation from natural language prompts. In the design phase, the framework achieves <3.2% prediction error for Power Usage Effectiveness (PUE). Compared to expert-built pure-physics models, it improves operational-phase twin accuracy by 37%, significantly accelerating automated deployment of high-fidelity, verifiable digital twins and closing the loop for intelligent operations and maintenance.

Technology Category

• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Machine Learning: Efficient ML / Green AI

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

The explosion in artificial intelligence (AI) applications is pushing the development of AI-dedicated data centers (AIDCs), creating management challenges that traditional methods and standalone AI solutions struggle to address. While digital twins are beneficial for AI-based design validation and operational optimization, current AI methods for their creation face limitations. Specifically, physical AI (PhyAI) aims to capture the underlying physical laws, which demands extensive, case-specific customization, and generative AI (GenAI) can produce inaccurate or hallucinated results. We propose Fusion Intelligence, a novel framework synergizing GenAI's automation with PhyAI's domain grounding. In this dual-agent collaboration, GenAI interprets natural language prompts to generate tokenized AIDC digital twins. Subsequently, PhyAI optimizes these generated twins by enforcing physical constraints and assimilating real-time data. Case studies demonstrate the advantages of our framework in automating the creation and validation of AIDC digital twins. These twins deliver predictive analytics to support power usage effectiveness (PUE) optimization in the design stage. With operational data collected, the digital twin accuracy is further improved compared with pure physics-based models developed by human experts. Fusion Intelligence offers a promising pathway to accelerate digital transformation. It enables more reliable and efficient AI-driven digital transformation for a broad range of mission-critical infrastructures.

Problem

• Managing AI-dedicated data centers with traditional methods is challenging
• Current AI methods for digital twin creation face accuracy limitations
• Synergizing GenAI and PhyAI to automate and optimize digital twins

Innovation

• Synergizes GenAI automation with PhyAI grounding
• Generates tokenized AIDC digital twins
• Optimizes twins via physical constraints