Moldable Task Scheduling in Dynamic SMP Clusters with Communication on the Fly

The paper concerns task graph scheduling in parallel programs using the concept of moldable computational tasks for a parallel architecture based on dynamic SMP processor clusters with data transmissions on the fly. The presented algorithm for scheduling parallel program graphs decomposes an initial program graph to sub-graphs, which fulfill the definition of a moldable task. So identified moldable tasks are then scheduled using an algorithm with warranted schedule length.

[1]  Marek Tudruj,et al.  Communication on the fly and program execution control in a system of dynamically configurable SMP clusters , 2003, Eleventh Euromicro Conference on Parallel, Distributed and Network-Based Processing, 2003. Proceedings..

[2]  Lukasz Masko.  Atomic operations for task scheduling for systems based on communication on-the-fly between SMP clusters , 2003, Second International Symposium on Parallel and Distributed Computing, 2003. Proceedings..

[3]  Lukasz Masko.  Program graph scheduling for dynamic SMP clusters with communication on the fly , 2004, Third International Symposium on Parallel and Distributed Computing/Third International Workshop on Algorithms, Models and Tools for Parallel Computing on Heterogeneous Networks.

[4]  Gerhard J. Woeginger,et al.  Approximation Algorithms for Scheduling Malleable Tasks under Precedence Constraints , 2001, ESA.

[5]  Andrei Tchernykh,et al.  Online Scheduling of Multiprocessor Jobs with Idle Regulation , 2003, PPAM.

[6]  Denis Trystram,et al.  Scheduling parallel applications using malleable tasks on clusters , 2001, Proceedings 15th International Parallel and Distributed Processing Symposium. IPDPS 2001.

[7]  Denis Trystram,et al.  A new clustering algorithm for large communication delays , 2002, Proceedings 16th International Parallel and Distributed Processing Symposium.