Comparative analysis of software optimization methods in context of branch predication on GPUs
General Purpose computing for Graphical Processing Units (GPGPU) technology is a powerful tool for offloading parallel data processing tasks to Graphical Processing Units (GPUs). This technology finds its use in variety of domains – from science and commerce to hobbyists. GPU-run general-purpose pro...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | Russian |
| Published: |
MIREA - Russian Technological University
2021-12-01
|
| Series: | Российский технологический журнал |
| Subjects: | |
| Online Access: | https://www.rtj-mirea.ru/jour/article/view/394 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849249584093593600 |
|---|---|
| author | I. Yu. Sesin R. G. Bolbakov |
| author_facet | I. Yu. Sesin R. G. Bolbakov |
| author_sort | I. Yu. Sesin |
| collection | DOAJ |
| description | General Purpose computing for Graphical Processing Units (GPGPU) technology is a powerful tool for offloading parallel data processing tasks to Graphical Processing Units (GPUs). This technology finds its use in variety of domains – from science and commerce to hobbyists. GPU-run general-purpose programs will inevitably run into performance issues stemming from code branch predication. Code predication is a GPU feature that makes both conditional branches execute, masking the results of incorrect branch. This leads to considerable performance losses for GPU programs that have large amounts of code hidden away behind conditional operators. This paper focuses on the analysis of existing approaches to improving software performance in the context of relieving the aforementioned performance loss. Description of said approaches is provided, along with their upsides, downsides and extents of their applicability and whether they address the outlined problem. Covered approaches include: optimizing compilers, JIT-compilation, branch predictor, speculative execution, adaptive optimization, run-time algorithm specialization, profile-guided optimization. It is shown that the aforementioned methods are mostly catered to CPU-specific issues and are generally not applicable, as far as branch-predication performance loss is concerned. Lastly, we outline the need for a separate performance improving approach, addressing specifics of branch predication and GPGPU workflow. |
| format | Article |
| id | doaj-art-f638ee8a20ff417e9dffb1dc46896f3d |
| institution | Kabale University |
| issn | 2782-3210 2500-316X |
| language | Russian |
| publishDate | 2021-12-01 |
| publisher | MIREA - Russian Technological University |
| record_format | Article |
| series | Российский технологический журнал |
| spelling | doaj-art-f638ee8a20ff417e9dffb1dc46896f3d2025-08-20T03:57:31ZrusMIREA - Russian Technological UniversityРоссийский технологический журнал2782-32102500-316X2021-12-019671510.32362/2500-316X-2021-9-6-7-15287Comparative analysis of software optimization methods in context of branch predication on GPUsI. Yu. Sesin0R. G. Bolbakov1MIREA – Russian Technological UniversityMIREA – Russian Technological UniversityGeneral Purpose computing for Graphical Processing Units (GPGPU) technology is a powerful tool for offloading parallel data processing tasks to Graphical Processing Units (GPUs). This technology finds its use in variety of domains – from science and commerce to hobbyists. GPU-run general-purpose programs will inevitably run into performance issues stemming from code branch predication. Code predication is a GPU feature that makes both conditional branches execute, masking the results of incorrect branch. This leads to considerable performance losses for GPU programs that have large amounts of code hidden away behind conditional operators. This paper focuses on the analysis of existing approaches to improving software performance in the context of relieving the aforementioned performance loss. Description of said approaches is provided, along with their upsides, downsides and extents of their applicability and whether they address the outlined problem. Covered approaches include: optimizing compilers, JIT-compilation, branch predictor, speculative execution, adaptive optimization, run-time algorithm specialization, profile-guided optimization. It is shown that the aforementioned methods are mostly catered to CPU-specific issues and are generally not applicable, as far as branch-predication performance loss is concerned. Lastly, we outline the need for a separate performance improving approach, addressing specifics of branch predication and GPGPU workflow.https://www.rtj-mirea.ru/jour/article/view/394general-purpose computing for graphical processing unitsoptimizing compilerspredication |
| spellingShingle | I. Yu. Sesin R. G. Bolbakov Comparative analysis of software optimization methods in context of branch predication on GPUs Российский технологический журнал general-purpose computing for graphical processing units optimizing compilers predication |
| title | Comparative analysis of software optimization methods in context of branch predication on GPUs |
| title_full | Comparative analysis of software optimization methods in context of branch predication on GPUs |
| title_fullStr | Comparative analysis of software optimization methods in context of branch predication on GPUs |
| title_full_unstemmed | Comparative analysis of software optimization methods in context of branch predication on GPUs |
| title_short | Comparative analysis of software optimization methods in context of branch predication on GPUs |
| title_sort | comparative analysis of software optimization methods in context of branch predication on gpus |
| topic | general-purpose computing for graphical processing units optimizing compilers predication |
| url | https://www.rtj-mirea.ru/jour/article/view/394 |
| work_keys_str_mv | AT iyusesin comparativeanalysisofsoftwareoptimizationmethodsincontextofbranchpredicationongpus AT rgbolbakov comparativeanalysisofsoftwareoptimizationmethodsincontextofbranchpredicationongpus |