Short-Term Load Forecasting for AI-Data Center

Mariam Mughees, Yuzhuo Li, Yize Chen, and Yunwei Ryan Li 1 · 2025

Read Short-Term Load Forecasting for AI-Data Center by Mariam Mughees, Yuzhuo Li, Yize Chen, and Yunwei Ryan Li 1 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-07756.txt.

Description

Mariam Mughees and co-authors built a forecasting workflow in 2025 for a new kind of power customer: the AI data center. Unlike steady factory loads, a hall of graphics processors can surge or stall within moments as training runs start, stop, and answer inference traffic. This stresses both the facility and the utility feeding it. Modeling that actions from first rules is impractical because the workloads are black boxes.

Instead the authors learn the pattern from data. Their workflow feeds logged consumption to sequence-learning methods, recurrent networks and 1D convolutional networks among them. And predicts the data center's short-term power draw, including at the level of its graphics processor fleet. Checked on real readings, the method reached decent accuracy, enough, they argue, for operators and utilities to allocate power dynamically as computing demand swings.

Catalog record

CREATORMariam Mughees, Yuzhuo Li, Yize Chen, and Yunwei Ryan Li 1
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE17 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-07756

Browse the full library · About the Tower of Babel