paperbot · PL 论文追踪

全部论文 待读 收藏

P2 Monitoring and Diagnosing Software Energy

软件能耗的监控与诊断
FSE Proceedings of the 34th ACM International Conference on the Foundations of Software Engineering2026匹配度 2/10 · 该论文是软件能耗的实时监控与热点诊断教程,属于动态性能/能耗测量工具方向,与用户关注的静态分析、LLM/Agent 交叉、浮点程序分析和贝叶斯程序分析均无直接关联。仅在“源码级细粒度诊断工具”和工业工具落地这一兴趣点上与用户有微弱重合。引用 0 inbox
Adel Noureddine
crossref: 2026-09-12dblp: 2026-09-12

AI 速览

一句话:本教程介绍实时软件能耗监控的主要方法及跨平台工具,帮助开发者从硬件到源码诊断能耗热点。

构建能效软件系统需要测量硬件与软件组件的能耗,包括源码方法、函数等细粒度构建块。本动手教程围绕实时能耗监控,介绍实现软件能耗监控的主要方法,并展示现代工具如何跨平台、跨操作系统,从硬件组件到源码层面采集能耗数据。此类细粒度且通用的监控可帮助开发者理解软件能耗,并诊断应用中的能耗热点。教程侧重于方法梳理与工具实践,而非提出新算法或实验评估。

问题
如何对软硬件能耗进行实时、细粒度且跨平台的监控与测量,以支持能效软件构建和能耗热点定位。
方法
以动手教程形式梳理实时能耗监控的主要方法,并演示覆盖多平台/操作系统、从硬件到源码的现代能耗监控工具。
评估
摘要未报告定量实验评估;展示方式为教程中的工具使用与实践演示。
原文摘要(Abstract)

Building energy efficient software systems requires monitoring and measuring the energy consumption of both hardware and software components, including individual building block such as source code methods and functions. In this hands-on tutorial, we present the main approaches in achieving real-time energy monitoring, and using modern tools to monitor software energy across platforms and operating systems, and from hardware components to source code monitoring. Such fine-grained and universal monitoring helps developers understand software energy and diagnose energy hotspots in their applications.

链接与引用

DOI 原文 · DBLP

BibTeX
@inproceedings{Noureddine26,
  title = {Monitoring and Diagnosing Software Energy},
  author = {Adel Noureddine},
  booktitle = {FSE},
  year = {2026},
  doi = {10.1145/3803437.3804901}
}