From cTuning.org
| Line 8: | Line 8: | ||
Collective Benchmark (cBench) is a collection of open-source sequential programs (working to add parallel applications) with multiple datasets (former MiDataSets) assembled by the community to enable realistic benchmarking and research on program and architecture optimization. The source code of individual programs is simplified to ease portability. All the benchmarks now include scripts to be used with [[CTools:CCC|Continuous Collective Compilation Framework]] to perform automatic optimizations (iterative compilation) using GCC, LLVM, GCC4CIL, Open64, PathScale, Intel and other compilers and a wide range of architectures. This collection of programs can later be used to create specialized benchmarks. | Collective Benchmark (cBench) is a collection of open-source sequential programs (working to add parallel applications) with multiple datasets (former MiDataSets) assembled by the community to enable realistic benchmarking and research on program and architecture optimization. The source code of individual programs is simplified to ease portability. All the benchmarks now include scripts to be used with [[CTools:CCC|Continuous Collective Compilation Framework]] to perform automatic optimizations (iterative compilation) using GCC, LLVM, GCC4CIL, Open64, PathScale, Intel and other compilers and a wide range of architectures. This collection of programs can later be used to create specialized benchmarks. | ||
| - | |||
| - | |||
| - | |||
| - | |||
| - | |||
| - | |||
cBench is an evolving project driven by the community demands - you are welcome to [[Join|join]] the project, [[CTools:CBench:Projects|extend or add]] benchmarks and datasets and provide performance analysis, leave [[CTools:CBench:Feedback|feedback]] and add your optimization data to the [[CDatabase|Collective Optimization Database]] to help the community optimize their programs. You can also communicate with cTuning community through our [[Community|mailing lists]]. | cBench is an evolving project driven by the community demands - you are welcome to [[Join|join]] the project, [[CTools:CBench:Projects|extend or add]] benchmarks and datasets and provide performance analysis, leave [[CTools:CBench:Feedback|feedback]] and add your optimization data to the [[CDatabase|Collective Optimization Database]] to help the community optimize their programs. You can also communicate with cTuning community through our [[Community|mailing lists]]. | ||
Revision as of 15:34, 5 September 2011
 |
Collective Benchmark |
| Enabling realistic benchmarking and optimization |
Web shortcut: http://cTuning.org/cbench
Navigation: cTuning.org > CTools
NOTES:
- Reference publications about cTuning.org long-term vision: GCC Summit'09, ACM TACO'10 journal and IJPP'11 journal.
Collective Benchmark (cBench) is a collection of open-source sequential programs (working to add parallel applications) with multiple datasets (former MiDataSets) assembled by the community to enable realistic benchmarking and research on program and architecture optimization. The source code of individual programs is simplified to ease portability. All the benchmarks now include scripts to be used with Continuous Collective Compilation Framework to perform automatic optimizations (iterative compilation) using GCC, LLVM, GCC4CIL, Open64, PathScale, Intel and other compilers and a wide range of architectures. This collection of programs can later be used to create specialized benchmarks.
cBench is an evolving project driven by the community demands - you are welcome to join the project, extend or add benchmarks and datasets and provide performance analysis, leave feedback and add your optimization data to the Collective Optimization Database to help the community optimize their programs. You can also communicate with cTuning community through our mailing lists.
- Projects to extend cBench (development plan and suggestions) - > 5 projects
- Current development:
cBench is an open collaborative community-driven project. You are welcome to join us and participate in discussions, developments or provide feedback and suggestions to extend cBench or add new benchmarks/datasets.