Energy Flow Graph: Modeling Software Energy Consumption

2026-03-17

AI Summary

This work addresses a critical limitation in existing approaches that treat software energy consumption as a monolithic, deterministic property, thereby overlooking substantial variations across different execution paths. To overcome this, the authors propose the Energy Flow Graph (EFG) model, which formally captures energy dependencies among program execution paths by representing software as a state-transition system annotated with state- and transition-level energy costs. Integrating static path analysis with a multiplicative cascade prediction model, the framework enables systematic green software optimization. Empirical evaluation across 3.5 million executions identified 15.6% of cases exhibiting high energy variance; structural optimizations achieved up to a 705× reduction in energy consumption. Furthermore, within AI pipelines, the method accurately predicts the optimal configuration among 4.2 million combinations with only 22 measurements and a prediction error of just 5.1%.

Technology Category

• Machine Learning: Efficient ML / Green AI
• Search and Optimization: Sampling/Simulation-based Search
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

The growing energy demands of computational systems necessitate a fundamental shift from performance-centric design to one that treats energy consumption as one of the primary design considerations. Current approaches treat energy consumption as an aggregate, deterministic property, overlooking the path-dependent nature of computation, where different execution paths through the same software consume dramatically different energy. We introduce the Energy Flow Graph (EFG), a formal model that represents computational processes as state-transition systems with energy costs for both states and transitions. EFG enables various applications in software engineering, including static analysis of energy-optimal execution paths and a multiplicative cascade model that predicts combined optimization effects without exhaustive testing. Our early experiments demonstrate EFG's versatility across domains: in software programs validated through 3.5 million executions, 15.6% of solutions exhibit high path-dependent variance (CV $>$ 0.1), while structural optimization reveals up to 705$\times$ energy reduction. In AI pipelines, the cascade model predicts optimization combinations within 5.1% error, enabling selection from 4.2 million possibilities using only 22 measurements. The EFG transforms energy optimization from trial-and-error to systematic analysis, providing a foundation for green software engineering across computational domains.

Problem

• energy consumption
• path-dependent
• software energy modeling
• execution paths
• computational energy

Innovation

• Energy Flow Graph
• path-dependent energy consumption
• static energy analysis
• multiplicative cascade model
• green software engineering