TY - GEN
T1 - Architecture-aware approximate computing
AU - Karakoy, Mustafa
AU - Kislal, Orhan
AU - Tang, Xulong
AU - Kandemir, Mahmut
AU - Arunachalam, Meenakshi
PY - 2019/6/20
Y1 - 2019/6/20
N2 - Observing that many application programs from different domains can live with less-than-perfect accuracy, existing techniques try to trade off program output accuracy with performance-energy savings. While these works provide point solutions, they leave three critical questions regarding approximate computing unanswered: (i) what is the maximum potential of skipping (i.e., not performing) data accesses under a given inaccuracy bound?; (ii) can we identify the data accesses to drop randomly, or is being architecture aware critical?; and (iii) do two executions that skip the same number of data accesses always result in the same output quality (error)? This paper first provides answers to these questions using ten multithreaded workloads, and then presents a program slicing-based approach that identifies the set of data accesses to drop. Results indicate 8.8% performance improvement and 13.7% energy saving are possible when we set the error bound to 2%, and the corresponding improvements jump to 15% and 25%, respectively, when the error bound is raised to 4%.
AB - Observing that many application programs from different domains can live with less-than-perfect accuracy, existing techniques try to trade off program output accuracy with performance-energy savings. While these works provide point solutions, they leave three critical questions regarding approximate computing unanswered: (i) what is the maximum potential of skipping (i.e., not performing) data accesses under a given inaccuracy bound?; (ii) can we identify the data accesses to drop randomly, or is being architecture aware critical?; and (iii) do two executions that skip the same number of data accesses always result in the same output quality (error)? This paper first provides answers to these questions using ten multithreaded workloads, and then presents a program slicing-based approach that identifies the set of data accesses to drop. Results indicate 8.8% performance improvement and 13.7% energy saving are possible when we set the error bound to 2%, and the corresponding improvements jump to 15% and 25%, respectively, when the error bound is raised to 4%.
UR - http://www.scopus.com/inward/record.url?scp=85069210662&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85069210662&partnerID=8YFLogxK
U2 - 10.1145/3309697.3331508
DO - 10.1145/3309697.3331508
M3 - Conference contribution
T3 - SIGMETRICS Performance 2019 - Abstracts of the 2019 SIGMETRICS/Performance Joint International Conference on Measurement and Modeling of Computer Systems
SP - 23
EP - 24
BT - SIGMETRICS Performance 2019 - Abstracts of the 2019 SIGMETRICS/Performance Joint International Conference on Measurement and Modeling of Computer Systems
PB - Association for Computing Machinery, Inc
T2 - 14th Joint Conference of International Conference on Measurement and Modeling of Computer Systems, SIGMETRICS 2019 and IFIP Performance Conference 2019, SIGMETRICS/Performance 2019
Y2 - 24 June 2019 through 28 June 2019
ER -