TY - GEN
T1 - Multi objective optimization of HPC kernels for performance, power, and energy
AU - Balaprakash, Prasanna
AU - Tiwari, Ananta
AU - Wild, Stefan M.
N1 - Publisher Copyright:
© Springer International Publishing Switzerland 2014.
PY - 2014
Y1 - 2014
N2 - Code optimization in the high-performance computing realm has traditionally focused on reducing execution time. The problem, in mathematical terms, has been expressed as a single objective optimization problem. The expected concerns of next-generation systems, however, demand a more detailed analysis of the interplay among execution time and other metrics. Metrics such as power, performance, energy, and resiliency may all be targeted together and traded against one another. We present a multi objective formulation of the code optimization problem. Our proposed framework helps one explore potential tradeoffs among multiple objectives and provides a significantly richer analysis than can be achieved by treating additional metrics as hard constraints. We empirically examine a variety of metrics, architectures, and code optimization decisions and provide evidence that such tradeoffs exist in practice.
AB - Code optimization in the high-performance computing realm has traditionally focused on reducing execution time. The problem, in mathematical terms, has been expressed as a single objective optimization problem. The expected concerns of next-generation systems, however, demand a more detailed analysis of the interplay among execution time and other metrics. Metrics such as power, performance, energy, and resiliency may all be targeted together and traded against one another. We present a multi objective formulation of the code optimization problem. Our proposed framework helps one explore potential tradeoffs among multiple objectives and provides a significantly richer analysis than can be achieved by treating additional metrics as hard constraints. We empirically examine a variety of metrics, architectures, and code optimization decisions and provide evidence that such tradeoffs exist in practice.
UR - http://www.scopus.com/inward/record.url?scp=84908662389&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-10214-6_12
DO - 10.1007/978-3-319-10214-6_12
M3 - Conference contribution
AN - SCOPUS:84908662389
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 239
EP - 260
BT - High Performance Computing Systems
A2 - Jarvis, Stephen A.
A2 - Wright, Steven A.
A2 - Hammond, Simon D.
PB - Springer Verlag
T2 - 4th International Workshop on Performance Modeling, Benchmarking and Simulation of High-Performance Computing Systems, PMBS 2013
Y2 - 18 November 2013 through 18 November 2013
ER -