BigPower: Hierarchical Source-Level Module Power Estimation for CPUs with Large Language Models
Researchers have developed BigPower, a new method for estimating power consumption in CPU modules during the design phase. This approach utilizes large language models to analyze source-level design information, including architectural hierarchy, module connectivity, and workload context. BigPower aims to provide a more efficient alternative to traditional simulation-based power estimation, as demonstrated by its successful application to the XiangShan processor family. AI
IMPACT This research could lead to more efficient CPU design by enabling faster and more granular power consumption analysis.