Abstract
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 that must be satisfied at run time. We have found that for nondeterministic programs, some of these actual can be satisfied with alternative data that can be generated in parallel, therefore boosting the program's TLP. We show how these (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%.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
