Computer Science > Distributed, Parallel, and Cluster Computing
[Submitted on 12 Jan 2016 (this version), latest version 22 Sep 2017 (v2)]
Title:Inferring Energy Bounds Statically by Evolutionary Analysis of Basic Blocks
View PDFAbstract:We are currently witnessing an increasing number of energy-bound devices, including in some cases mission critical systems, for which there is a need to optimize their energy consumption and verify that they will perform their function within the available energy budget. In this work we propose a novel parametric approach to estimating tight energy bounds (both upper and lower) that are practical for energy verification and optimization applications in embedded systems. Our approach consists in dividing a program into basic ("branchless") blocks, establishing the maximal (resp. minimal) energy consumption for each block using an evolutionary algorithm, and combining the obtained values according to the program control flow, using static analysis, to produce energy bound functions. Such functions depend on input data sizes, and return upper or lower bounds on the energy consumption of the program for any given set of input values of those sizes, without running the program. The approach has been tested on XMOS chips, but is general enough to be applied to any microprocessor and programming language. Our experimental results show that the bounds obtained by our prototype tool can be tight while remaining on the safe side of budgets in practice.
Submission history
From: Pedro Lopez-Garcia [view email][v1] Tue, 12 Jan 2016 10:52:48 UTC (219 KB)
[v2] Fri, 22 Sep 2017 07:46:56 UTC (203 KB)
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.