Abstract-In the era of multicore systems, it is expected that the number of cores that can be integrated on a single chip will be 3-digit. The key to utilize such a huge computational power is to extract the very fine parallelism in the user program. This is nontrivial for the average programmer, and becomes very hard as the number of potential parallel instances increases. Task-based programming models such as OmpSs are promising, since they handle the detection of dependencies and synchronization for the programmer. However, state-of-the-art research shows that task management is not cheap, and introduces a significant overhead that limits the scalability of OmpSs. Nexus# is a hardware accelerator for the OmpSs runtime system, which dynamically monitors dependencies between tasks. It is fully synthesizable in VHDL, and has a distributed task graph model to achieve the best scalability. Supporting tasks with arbitrary number of parameters and any dependency pattern, Nexus# achieves better performance than Nanos, the official OmpSs runtime system, and scales well for the H264dec benchmark with very fine grained tasks, among other benchmarks from the Starbench suite.

T am e r D a l l o u , B e n J u u r l i n k H a r dw a r e -b a s e d t a s k d e p e n d e n c y r e s o l u t i o n f o r t h e S t a r S s p r o g r amm i n g m o d e l C on f e r en c e ob j e c t , P o s tp r in t v e r s i on T h i s v e r s i o n i s a v a i l a b l e a t h t t p : / / d x . d o i . o r g / 1 0 . 1 4 2 7 9 / d e p o s i t o n c e -5 7 8 1 .Sugg e s t ed C i t a t i on D a l l o u , T am e r ; J u u r l i n k , B e n : H a r dw a r e -b a s e d t a s k d e p e n d e n c y r e s o l u t i o n f o r t h e S t a r S s p r o g r amm i n g m o d e l . -I n : 2 0 1 2 4 1 s t I n t e r n a t i o n a l C o n f e r e n c e o n P a r a l l e l P r o c e s s i n g W o r k s h o p s : I C P PW . -N ew Y o r k , N Y [ u . a . ] : I E E E , 2 0 1 2 . -I S BN : 9 7 8 -1 -4 6 7 3 -2 5 0 9 -7 . -p p . 3 6 7 -3 7 4 . -DO I : 1 0 . 1 1 0 9 / I C P PW . 2 0 1 2 . 5 3 . ( P o s t p r i n t v e r s i o n i s c i t e d , p a g e n um b e r s d i f f e r . ) T e rm s o f U s e © © 2 0 1 2 I E E E . P e r s o n a l u s e o f t h i s m a t e r i a l i s p e rm i t t e d . P e rm i s s i o n f r om I E E E m u s t b e o b t a i n e d f o r a l l o t h e r u s e s , i n a n y c u r r e n t o r f u t u r e m e d i a , i n c l u d i n g r e p r i n t i n g / r e p u b l i s h i n g t h i s m a t e r i a l f o r a d v e r t i s i n g o r p r om o t i o n a l p u r p o s e s , c r e a t i n g n ew c o l l e c t i v e w o r k s , f o r r e s a l e o r r e d i s t r i b u t i o n t o s e r v e r s o r l i s t s , o r r e u s e o f a n y c o p y r i g h t e d c om p o n e n t o f t h i s w o r k i n o t h e r w o r k s .

Abstract-Task-based parallel programming models with explicit data dependencies, such as OmpSs, are gaining popularity, due to the ease of describing parallel algorithms with complex and irregular dependency patterns. These advantages, however, come at a steep cost of runtime overhead incurred by dynamic dependency resolution. Hardware support for task management has been proposed in previous work as a possible solution. We present VSs, a runtime library for the OmpSs programming model that integrates the Nexus++ hardware task manager, and evaluate the performance of the VSs-Nexus++ system. Experimental results show that applications with fine-grain tasks can achieve speedups of up to 3.4×, while applications optimized for current runtimes attain 1.3×. Providing support for hardware task managers in runtime libraries is therefore a viable approach to improve the performance of OmpSs applications.

No abstract

No abstract