Fix  the  code.  Don’t  tweak  the  hardware:  A  newcompiler  approach  to  Voltage-Frequency  scaling

Jimborean, Alexandra; Koukos, Konstantinos; Spiliopoulos, Vasileios; Black-Schaffer, David; Kasiras, Stefanos

Publication:
Fix the code. Don’t tweak the hardware: A newcompiler approach to Voltage-Frequency scaling

dc.contributor.author	Jimborean, Alexandra
dc.contributor.author	Koukos, Konstantinos
dc.contributor.author	Spiliopoulos, Vasileios
dc.contributor.author	Black-Schaffer, David
dc.contributor.author	Kasiras, Stefanos
dc.contributor.department	Ingeniería y Tecnología de Computadores
dc.date.accessioned	2024-02-01T08:35:30Z
dc.date.available	2024-02-01T08:35:30Z
dc.date.created	2014
dc.description.abstract	Traditional compiler approaches to optimize power effi-ciency aim to adjust voltage and frequency at runtime tomatch the code characteristics to the hardware (e.g., run-ning memory-bound phases at a lower frequency). However,such approaches are constrained by three factors: (i) voltage-frequency transitions are too slow to be applied at instructiongranularity, (ii) larger code regions are seldom unequivocallymemory- or compute-bound, and, (iii) the available voltagescaling range for future technologies is rapidly shrinking.These factors necessitate new approaches to address power-efficiency at the code-generation level. This paper proposesone such approach to automatically generate power-efficientcode using a decoupled access/execute (DAE) model.In DAE a program is split into tasks, where each task con-sists of two sufficiently coarse-grained phases to enable ef-fective Dynamic Voltage Frequency Scaling (DVFS): (i) theaccess-phase for data prefetch (heavily memory-bound), and(ii) theexecute-phase that performs the actual computation(heavily compute-bound). Our contribution is to provide acompiler methodology to automatically generate the access-phases for a task-based programming system. Our approachis capable of handling both affine (through a polyhedral anal-ysis) and non-affine codes (through optimized task skele-tons). Our evaluation shows that the automatically gener-ated versions improve EDP by 25% on average comparedto a coupled execution, without any performance degrada-tion, and surpasses the EDP savings of the correspondinghand-crafted tasks by 5%	es
dc.embargo.terms	1-ene-2999
dc.format	application/pdf	es
dc.format.extent	11	es
dc.identifier.citation	IEEE/ACM Transactions on Networking, February 15-19 2014: 262 - 272
dc.identifier.doi	http://dx.doi.org/10.1145/2544137.2544161
dc.identifier.issn	1558-2566
dc.identifier.uri	http://hdl.handle.net/10201/138368
dc.language	eng	es
dc.publisher	IEEE/ACM	es
dc.relation	We thank Vincent Loechner for his insights on the polyhedral model and his valuable help. This work is supported, in part, by the Swedish Research Council UPMARC Linnaeus Centre and the EU Project: LPGPU FP7-ICT-288653.	es
dc.relation.ispartof	12th Annual IEEE/ACM International Symposium on Code Generation and Optimization 2014	es
dc.relation.publisherversion	https://dl.acm.org/citation.cfm?id=2544161	es
dc.rights	Attribution-NonCommercial-NoDerivatives 4.0 Internacional	*
dc.rights.accessRights	info:eu-repo/semantics/restrictedAccess
dc.subject	Compiler Optimizations	es
dc.subject	Polyhedral Model	es
dc.subject	Task-Based Execution	es
dc.subject	Execution	es
dc.subject	Performance	es
dc.subject	Energy	es
dc.subject	DVFS	es
dc.title	Fix the code. Don’t tweak the hardware: A newcompiler approach to Voltage-Frequency scaling	es
dc.type	info:eu-repo/semantics/article	es
dspace.entity.type	Publication	es