Evaluating GPP Predictors for Software Based Waveform Performance

  • William “BJ” Blair ANDRO Computational Solutions
  • Ashwin Amanna ANDRO Computational Solutions
  • Timothy Reichert ANDRO Computational Solutions
  • Michael Gudaitis Air Force Research Laboratory

Abstract

Making full use of a computer’s capabilities today is a challenging task due to increased hardware and software complexity, requiring the use of multithreading, SIMD intrinsics, and overclocking to squeeze as much performance out of a system as possible. A challenge is predicting how a software-based waveforms will perform based on published benchmarks on a general purpose processor (GPP) of interest and where the key limiters exist. This is valuable insight to determine implementation and optimization strategies for software-based waveforms. This paper attempts to identify key indicators of modern GPP performance for usage with waveform software, using LDPC and DVB-S2 waveform benchmarks on two consumer grade desktops. We find a correlation in software performance between GPP or memory reliance and GPP clock speed and cache, as well as the importance of system tuning and overclocking.

Published
2023-09-13
How to Cite
BLAIR, William “BJ” et al. Evaluating GPP Predictors for Software Based Waveform Performance. Proceedings of the GNU Radio Conference, [S.l.], v. 8, n. 1, sep. 2023. Available at: <https://pubs.gnuradio.org/index.php/grcon/article/view/135>. Date accessed: 21 apr. 2024.