The Unseen AI Disruptions for Power Grids: LLM-Induced Transients

TL;DR

This paper investigates the disruptive transient power consumption behaviors of large language models (LLMs), highlighting their potential threats to power grid stability and emphasizing the need for interdisciplinary solutions for sustainable AI infrastructure.

Contribution

It introduces a comprehensive analysis of AI transient power dynamics, develops mathematical models, and discusses challenges and opportunities for integrating AI loads into power grids.

Findings

AI workloads cause sharp power surges and dips.

AI power consumption exhibits ultra-low inertia and high peak-idle ratios.

Potential threats to power grid reliability from AI transient behaviors.

Abstract

Recent breakthroughs of large language models (LLMs) have exhibited superior capability across major industries and stimulated multi-hundred-billion-dollar investment in AI-centric data centers in the next 3-5 years. This, in turn, bring the increasing concerns on sustainability and AI-related energy usage. However, there is a largely overlooked issue as challenging and critical as AI model and infrastructure efficiency: the disruptive dynamic power consumption behaviour. With fast, transient dynamics, AI infrastructure features ultra-low inertia, sharp power surge and dip, and a significant peak-idle power ratio. The power scale covers from several hundred watts to megawatts, even to gigawatts. These never-seen-before characteristics make AI a very unique load and pose threats to the power grid reliability and resilience. To reveal this hidden problem, this paper examines the scale of AI power consumption, analyzes AI transient behaviour in various scenarios, develops high-level mathematical models to depict AI workload behaviour and discusses the multifaceted challenges and opportunities they potentially bring to existing power grids. Observing the rapidly evolving machine learning (ML) and AI technologies, this work emphasizes the critical need for interdisciplinary approaches to ensure reliable and sustainable AI infrastructure development, and provides a starting point for researchers and practitioners to tackle such challenges.