Abstract
Models based on task graphs that operate on single-assignment data are attractive in several ways, but also require nuanced algorithms for scheduling and memory management for efficient execution. In this paper, we consider memory-efficient dynamic scheduling of task graphs, and present a novel approach for dynamically recycling the memory locations assigned to data items as they are produced by tasks.
- L. Gérard, A. Guatto, C. Pasteur, and M. Pouzet. A modular memory optimization for synchronous data-flow languages: application to arrays in a Lustre compiler. In LCTES, pages 51--60, 2012. Google Scholar
Digital Library
- M. C. Kurt, S. Krishnamoorthy, K. Agrawal, and G. Agrawal. Fault-tolerant dynamic task graph scheduling. In SC, pages 719--730, 2014. Google Scholar
Digital Library
- V. Raychev, M. Vechev, and M. Sridharan. Effective race detection for event-driven programs. In OOPSLA, pages 151--166, 2013. Google Scholar
Digital Library
- D. Sbirlea, Z. Budimlic, and V. Sarkar. Bounded memory scheduling of dynamic task graphs. In PACT, pages 343--356, 2014. Google Scholar
Digital Library
- D. Sbirlea, K. Knobe, and V. Sarkar. Folding of tagged single assignment values for memory-efficient parallelism. In Euro-Par, pages 601--613, 2012. Google Scholar
Digital Library
- P. Schnorf, M. Ganapathi, and J. L. Hennessy. Compile-time copy elimination. Softw., Pract. Exper., 23(11):1175--1200, 1993. Google Scholar
Digital Library
Index Terms
User-assisted storage reuse determination for dynamic task graphs
Recommendations
User-assisted storage reuse determination for dynamic task graphs
PPoPP '16: Proceedings of the 21st ACM SIGPLAN Symposium on Principles and Practice of Parallel ProgrammingModels based on task graphs that operate on single-assignment data are attractive in several ways, but also require nuanced algorithms for scheduling and memory management for efficient execution. In this paper, we consider memory-efficient dynamic ...
User-Assisted Store Recycling for Dynamic Task Graph Schedulers
The emergence of the multi-core era has led to increased interest in designing effective yet practical parallel programming models. Models based on task graphs that operate on single-assignment data are attractive in several ways. Notably, they can ...
Efficient dynamic heap allocation of scratch-pad memory
ISMM '08: Proceedings of the 7th international symposium on Memory managementAn increasing number of processor architectures support scratch-pad memory - software managed on-chip memory. Scratch-pad memory provides low latency data storage, like on-chip caches, but under explicit software control. The simple design and ...






Comments