OpenCL performance prediction using architecture-independent features

Beau Johnston, Gregory Falzon, Josh Milthorpe

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

OpenCL is an attractive programming model for heterogeneous high-performance computing systems, with wide support from hardware vendors and significant performance portability. To support efficient scheduling on HPC systems it is necessary to perform accurate performance predictions for OpenCL workloads on varied compute devices, which is challenging due to diverse computation, communication and memory access characteristics which result in varying performance between devices. The Architecture Independent Workload Characterization (AIWC) tool can be used to characterize OpenCL kernels according to a set of architecture-independent features. This work presents a methodology where AIWC features are used to form a model capable of predicting accelerator execution times. We used this methodology to predict execution times for a set of 37 computational kernels running on 15 different devices representing a broad range of CPU, GPU and MIC architectures. The predictions are highly accurate, differing from the measured experimental run-Times by an average of only 1.2%, and correspond to actual execution time mispredictions of 9 ps to 1 sec according to problem size. A previously unencountered code can be instrumented once and the AIWC metrics embedded in the kernel, to allow performance prediction across the full range of modelled devices. The results suggest that this methodology supports correct selection of the most appropriate device for a previously unen-countered code, which is highly relevant to the HPC scheduling setting.

Original languageEnglish
Title of host publicationProceedings - 2018 International Conference on High Performance Computing and Simulation, HPCS 2018
EditorsKhalid Zine-Dine, Waleed W. Smari
Place of PublicationPiscataway, NJ
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages561-569
Number of pages9
ISBN (Electronic)9781538678787
ISBN (Print)9781538678794
DOIs
Publication statusPublished - 29 Oct 2018
Externally publishedYes
Event16th International Conference on High Performance Computing and Simulation, HPCS 2018 - Orleans, France
Duration: 16 Jul 201820 Jul 2018

Publication series

NameProceedings - 2018 International Conference on High Performance Computing and Simulation, HPCS 2018

Conference

Conference16th International Conference on High Performance Computing and Simulation, HPCS 2018
CountryFrance
CityOrleans
Period16/07/1820/07/18

Keywords

  • Accelerator
  • HPC
  • Modelling
  • Prediction
  • Supercomputing
  • Workload characterization

Fingerprint Dive into the research topics of 'OpenCL performance prediction using architecture-independent features'. Together they form a unique fingerprint.

  • Cite this

    Johnston, B., Falzon, G., & Milthorpe, J. (2018). OpenCL performance prediction using architecture-independent features. In K. Zine-Dine, & W. W. Smari (Eds.), Proceedings - 2018 International Conference on High Performance Computing and Simulation, HPCS 2018 (pp. 561-569). [8514400] (Proceedings - 2018 International Conference on High Performance Computing and Simulation, HPCS 2018). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/HPCS.2018.00095