Mostra i principali dati dell'item
Towards automatic significance analysis for approximate computing
dc.creator | Vassiliadis V., Riehme J., Deussen J., Parasyris K., Antonopoulos C.D., Bellas N., Lalis S., Naumann U. | en |
dc.date.accessioned | 2023-01-31T10:29:13Z | |
dc.date.available | 2023-01-31T10:29:13Z | |
dc.date.issued | 2016 | |
dc.identifier | 10.1145/2854038.2854058 | |
dc.identifier.isbn | 9781450337786 | |
dc.identifier.uri | http://hdl.handle.net/11615/80496 | |
dc.description.abstract | Several applications may trade-off output quality for energy efficiency by computing only an approximation of their output. Current approaches to software-based approximate computing often require the programmer to specify parts of the code or data structures that can be approximated. A largely unaddressed challenge is how to automate the analysis of the significance of code for the output quality. To this end, we propose a methodology and toolset for automatic significance analysis. We use interval arithmetic and algorithmic differentiation in our profile-driven yet mathematical approach to evaluate the significance of input and intermediate variables for the output of a computation. Our methodology effectively matches decisions of a domain expert in significance characterization for a set of benchmarks, and in some cases offers new insights. Evaluation of the software infrastructure on a multicore x86 platform shows energy reduction (from 31% up to 91% with a mean of 56%) compared to fully accurate execution, with graceful quality degradation. © 2016 ACM. | en |
dc.language.iso | en | en |
dc.source | Proceedings of the 14th International Symposium on Code Generation and Optimization, CGO 2016 | en |
dc.source.uri | https://www.scopus.com/inward/record.uri?eid=2-s2.0-84968763937&doi=10.1145%2f2854038.2854058&partnerID=40&md5=7a13204e98e585929a503c77a3bb7a38 | |
dc.subject | Codes (symbols) | en |
dc.subject | Economic and social effects | en |
dc.subject | Energy efficiency | en |
dc.subject | Network components | en |
dc.subject | Algorithmic differentiations | en |
dc.subject | Approximate computing | en |
dc.subject | Energy reduction | en |
dc.subject | Interval arithmetic | en |
dc.subject | Mathematical approach | en |
dc.subject | Quality degradation | en |
dc.subject | Significance analysis | en |
dc.subject | Software infrastructure | en |
dc.subject | Quality control | en |
dc.subject | Association for Computing Machinery, Inc | en |
dc.title | Towards automatic significance analysis for approximate computing | en |
dc.type | conferenceItem | en |
Files in questo item
Files | Dimensione | Formato | Mostra |
---|---|---|---|
Nessun files in questo item. |