Tuning Performance via Metrics with Expectations

Ahmad Yasin, Avi Mendelson, Yosi Ben-Asher

Research output: Contribution to journalArticlepeer-review


Modern server systems employ many features that are difficult to exploit by software developers. This paper calls for a new performance optimization approach that uses designated metrics with expected optimal values. A key insight is that expected values of these metrics are essential in order to verify that no performance is wasted during incremental utilization of processor features. We define sample primary metrics for modern architectures and present three distinct techniques that help to determine their optimal values. Our preliminary results successfully provide 2\text{x}\hbox{-} 4\text{x}2x-4x extra speedup during tuning of commonly-used software optimizations on the matrix-multiply kernel. Additionally, our approach helped to identify counter-intuitive causes that hurt multicore scalability of an optimized deep-learning benchmark on a Cascade Lake server.

Original languageEnglish
Article number8714063
Pages (from-to)91-94
Number of pages4
JournalIEEE Computer Architecture Letters
Issue number2
StatePublished - 1 Jul 2019

Bibliographical note

Publisher Copyright:
© 2002-2011 IEEE.


  • Code tuning
  • SIMD processors
  • measurements
  • micro-architecture
  • multi-core/single-chip multiprocessors
  • optimization
  • performance analysis

ASJC Scopus subject areas

  • Hardware and Architecture


Dive into the research topics of 'Tuning Performance via Metrics with Expectations'. Together they form a unique fingerprint.

Cite this