Mapping and Synchronizing Streaming Applications on Cell Processors - HiPEAC 2009

Mapping and Synchronizing Streaming Applications on Cell Processors - HiPEAC 2009

Maik Nijhuis, Herbert Bos, Henri E. Bal, and Cédric Augonnet

Abstract. Developing streaming applications on heterogenous multi-processor architectures like the Cell is difficult. Currently, application developers need to know about hardware details to deal with issues like scheduling, memory man- agement and communication/synchronization. Worse, with multiple alternatives for communication available, developers spend significant time picking the most appropriate one. A poor choice often results in bad performance. With Cell- Space, we shield users from hardware details without compromising perfor- mance. Its runtime is based on an evaluation of the different communication primitives. In Cell-Space, developers specify a streaming application as a data flow graph of interacting components. Both task- and data-parallelism are eas- ily expressed and advanced features such as dynamic reconfiguration are fully supported. Beneath a simple interface we include a slew of optimizations not present in other Cell run time environments. We demonstrate the impact of these optimizations and show that Cell-Space applications can efficiently exploit the resources offered by the Cell.

Download

Bibtex

@InProceedings{NijBosBalAug09HiPEAC, author= {Maik Nijhuis and Herbert Bos and Henri Bal and C'edric Augonnet}, title = {{Mapping and Synchronizing Streaming Applications on Cell Processors}}, address = {Paphos, Cyprus}, month = JAN, year = {2009}, booktitle = {International Conference on High Performance Embedded Architectures & Compilers}, doi = {10.1007/978-3-540-92990-1_17}, url = {http://runtime.bordeaux.inria.fr/augonnet/papers/NijBosBalAug09HiPEAC.pdf} }