Adaptive configuration selection for power-constrained heterogeneous systems

Peter E. Bailey, David K. Lowenthal, Vignesh Ravi, Barry Rountree, Martin Schulz, Bronis R. De Supinski

Research output: Contribution to journalConference article

38 Scopus citations

Abstract

As power becomes an increasingly important design factor in high-end supercomputers, future systems will likely operate with power limitations significantly below their peak power specifications. These limitations will be enforced through a combination of software and hardware power policies, which will filter down from the system level to individual nodes. Hardware is already moving in this direction by providing power-capping interfaces to the user. The power/performance trade-off at the node level is critical in maximizing the performance of power-constrained cluster systems, but is also complex because of the many interacting architectural features and accelerators that comprise the hardware configuration of a node. The key to solving this challenge is an accurate power/performance model that will aid in selecting the right configuration from a large set of available configurations. In this paper, we present a novel approach to generate such a model offline using kernel clustering and multivariate linear regression. Our model requires only two iterations to select a configuration, which provides a significant advantage over exhaustive search-based strategies. We apply our model to predict power and performance for different applications using arbitrary configurations, and show that our model, when used with hardware frequency-limiting, selects configurations with significantly higher performance at a given power limit than those chosen by frequency-limiting alone. When applied to a set of 36 computational kernels from a range of applications, our model accurately predicts power and performance, it maintains 91% of optimal performance while meeting power constraints 88% of the time. When the model violates a power constraint, it exceeds the constraint by only 6% in the average case, while simultaneously achieving 54% more performance than an oracle.

Original languageEnglish (US)
Article number6957246
Pages (from-to)371-380
Number of pages10
JournalProceedings of the International Conference on Parallel Processing
Volume2014-November
Issue numberNovember
DOIs
StatePublished - Nov 13 2014
Event43rd International Conference on Parallel Processing, ICPP 2014 - Minneapolis, United States
Duration: Sep 9 2014Sep 12 2014

Keywords

  • GPU APU power performance modeling power-constrained

ASJC Scopus subject areas

  • Software
  • Mathematics(all)
  • Hardware and Architecture

Fingerprint Dive into the research topics of 'Adaptive configuration selection for power-constrained heterogeneous systems'. Together they form a unique fingerprint.

  • Cite this