Logo PTI Logo FedCSIS

Proceedings of the 21st Conference on Computer Science and Intelligence Systems (FedCSIS)

Annals of Computer Science and Information Systems, Volume 47

Assessing Performance Portability of Hybrid Video Encoders

,

DOI: http://dx.doi.org/10.15439/2026F3256

Citation: Maciej Okoń,

Full text

Abstract. While the Java Virtual Machine ensures functional code portability, achieving consistent execution efficiency across hardware generations remains a significant challenge for compute-intensive multimedia applications. This paper evaluates performance portability in a hybrid video encoder, focusing on the interaction between algorithmic complexity and parallelization granularity. We contrast a legacy fine-grained model, utilizing macroblock-level task scheduling, with a coarse-grained architecture based on Group of Pictures (GOP) parallelism. The study quantifies how these two paradigms respond to increasing core density and generational shifts in server-grade hardware. The evaluation utilizes three configurations: an intra-frame baseline, a hybrid model with integer-pixel motion estimation, and a sub-pixel hybrid model. Stability is assessed across four server generations using a framework that combines harmonic portability metrics with arithmetic efficiency averages and cascade visualizations. This approach identifies the exact points where task synchronization and memory latency limit hardware utilization in managed execution environments. The results demonstrate that the GOP-parallel implementation significantly enhances performance portability, achieving architectural portability metrics between 0.83 and 0.87. In contrast, the legacy macroblock-level model exhibits severe efficiency degradation on modern high-density processors, with portability scores dropping as low as 0.41. Visualization through efficiency cascades and platform fingerprints confirms that the coarse-grained GOP model effectively mitigates synchronization overheads and latency sensitivities. The study demonstrates that fine-grained parallelization strategies are insufficient for maintaining performance portability on modern many-core architectures within managed environments, whereas a coarse-grained GOP-level approach provides a more robust alternative for achieving consistent execution efficiency.

References

  1. B. Bylina and M. Okoń, “A Multithreaded Java-Based Video Encoder for Multicore Systems,” in Proc. 20th Conf. Computer Science and Intelligence Systems (FedCSIS), ACSIS, vol. 43, 2025, pp. 659–664. https://dx.doi.org/10.15439/2025F1886.
  2. S. Sankaraiah, L. H. Shuan, C. Eswaran, and J. Abdullah, “Performance Optimization of Video Coding Process on Multi-Core Platform Using GOP Level Parallelism,” International Journal of Parallel Programming, vol. 42, no. 6, pp. 931–947, Dec. 2014. https://dx.doi.org/10.1007/s10766-013-0267-4.
  3. T. Wiegand, G. J. Sullivan, G. Bjøntegaard, and A. Luthra, “Overview of the H.264/AVC Video coding standard,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 13, no. 7, pp. 560–576, 2003. https://dx.doi.org/10.1109/TCSVT.2003.815165.
  4. Oracle, “Concurrency Utilities,” Java Platform, Standard Edition 8 API Specification, 2014. [Online]. Available: https://docs.oracle.com/javase/8/docs/api/java/util/concurrent/packagesummary.html.
  5. G. J. Sullivan, J. Ohm, W. Han, and T. Wiegand,“Overview of the High Efficiency Video Coding (HEVC) Standard”, in IEEE Transactions on Circuits and Systems for Video Technology, vol. 22, no. 12, pp. 1649– 1668, Dec. 2012. https://dx.doi.org/10.1109/TCSVT.2012.2221191.
  6. I. E. G. Richardson, H.264 and MPEG-4 Video Compression: Video Coding for Next-generation Multimedia. Chichester, U.K.: Wiley, 2003.
  7. V. Sze, M. Budagavi, and G. J. Sullivan, High Efficiency Video Coding (HEVC): Algorithms and Architectures. Cham, Switzerland: Springer, 2014. https://dx.doi.org/10.1007/978-3-319-06895-6.
  8. S. J. Pennycook, J. D. Sewall, and V. W. Lee, “A Metric for Performance Portability,” arXiv preprint, 2016. [Online]. Available: https://doi.org/10. 48550/arXiv.1611.07409.
  9. A. Marowka, “Portability Efficiency Approach for Calculating Performance Portability,” arXiv preprint, Nov. 2024. [Online]. Available: https: //doi.org/10.48550/arXiv.2407.00232.
  10. J. Sewall, S. Pennycook, D. Jacobsen, T. Deakin, and S. McIntosh-Smith, “Interpreting and Visualizing Performance Portability Metrics,” in Proc. Int. Workshop Perform. Portability Product. HPC (P3HPC), Nov. 2020, pp. 14–24. https://dx.doi.org/10.1109/P3HPC51967.2020.00007.