An online learning methodology for performance modeling of graphics processors

Ujjwal Gupta, Manoj Babu, Raid Ayoub, Michael Kishinevsky, Francesco Paterna, Suat Gumussoy, Umit Ogras

Research output: Contribution to journalArticlepeer-review

8 Scopus citations


Approximately 18 percent of the 3.2 million smartphone applications rely on integrated graphics processing units (GPUs) to achieve competitive performance. Graphics performance, typically measured in frames per second, is a strong function of the GPU frequency, which in turn has a significant impact on mobile processor power consumption. Consequently, dynamic power management algorithms have to assess the performance sensitivity to the frequency accurately to choose the operating frequency of the GPU effectively. Since the impact of GPU frequency on performance varies rapidly over time, there is a need for online performance models that can adapt to varying workloads. This paper presents a light-weight adaptive runtime performance model that predicts the frame processing time of graphics workloads at runtime without apriori characterization. We employ this model to estimate the frame time sensitivity to the GPU frequency, i.e., the partial derivative of the frame time with respect to the GPU frequency. The proposed model does not rely on any parameter learned offline. Our experiments on commercial platforms with common GPU benchmarks show that the mean absolute percentage error in frame time and frame time sensitivity prediction are 4.2 and 6.7 percent, respectively.

Original languageEnglish (US)
Article number8365819
Pages (from-to)1677-1691
Number of pages15
JournalIEEE Transactions on Computers
Issue number12
StatePublished - Dec 1 2018


  • Integrated GPUs
  • RLS
  • frequency scaling
  • online learning
  • performance modeling
  • power management

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Hardware and Architecture
  • Computational Theory and Mathematics


Dive into the research topics of 'An online learning methodology for performance modeling of graphics processors'. Together they form a unique fingerprint.

Cite this