The demand for thread-level parallelism (TLP) is endless, especially on commodity processors, as TLP is essential for gaining performance. However, the TLP of today's programs is limited by dependences that must be satisfied at run time. We have found that for nondeterministic programs, some of these actual dependences can be satisfied with alternative data that can be generated in parallel, therefore boosting the program's TLP. We show how these dependences (which we call 'state dependences' because they are related to the program's state) can be exploited using algorithm-specific knowledge. To demonstrate the practicality of our technique, we implemented a system called April25th that incorporates the concept of 'state dependences'. This system boosts the performance of five nondeterministic, multi-threaded PARSEC benchmarks by 100.5%.
|Title of host publication
|Proceedings - 26th International Conference on Parallel Architectures and Compilation Techniques, PACT 2017
|Institute of Electrical and Electronics Engineers Inc.
|Number of pages
|Published - Oct 31 2017
|26th International Conference on Parallel Architectures and Compilation Techniques, PACT 2017 - Portland, United States
Duration: Sep 9 2017 → Sep 13 2017
|Parallel Architectures and Compilation Techniques - Conference Proceedings, PACT
|26th International Conference on Parallel Architectures and Compilation Techniques, PACT 2017
|9/9/17 → 9/13/17
ASJC Scopus subject areas
- Theoretical Computer Science
- Hardware and Architecture