|Ομιλητής||Αντώνης Χρονόπουλος, University of Texas at San Antonio Department of Computer Science|
|Τίτλος||Hierarchical Distributed Loop Self-Scheduling Schemes on Cluster and Cloud Systems|
|Ημερομηνία||Τρίτη 29/10/2013, ώρα 15:00|
|Χώρος||Αίθουσα Συνεδριάσεων, 4ος όροφος|
|Διεύθυνση||Γκλαβάνη 37, Βόλος|
Anthony T. Chronopoulos received his Ph.D. in Computer Science from the University of Illinois at Urbana-Champaign in 1987. He is currently a Professor in Computer Science at the University of Texas at San Antonio. He has published 50 journal and over 65 peer-reviewed conference proceedings publications in the areas of Distributed and Parallel Computing, High Performance Computing, Scientific Computing. He has been awarded 15 federal/state government research grants. His work is cited in over 1300 non-coauthors’ research articles with h-index=24. He is a senior member of IEEE (since 1997).
Y. Han, A. T. Chronopoulos, Distributed Loop Scheduling Schemes for Cloud Systems, Proceedings of the 27th IEEE International Parallel and Distributed Processing Symposium, High-Performance Grid and Cloud Computing Workshop, pp. 955-962, Boston, Massachusetts, USA, May 2013.
Y. Han, A. T. Chronopoulos, Scalable Loop Self-Scheduling Schemes Implemented on Large-Scale Clusters, Proceedings of the 27th IEEE International Parallel and Distributed Processing Symposium, Large-Scale Parallel Processing Workshop, pp. 1735-1742, Boston, Massachusetts, USA, May 2013.
Loops are the largest source of parallelism in many scientific applications. Parallelization of irregular loop applications is a challenging problem to achieve scalable performance on cluster and cloud systems. For cluster systems, previous research proposed an effective Master-Worker model on clusters for distributed self-scheduling schemes that apply to parallel loops with independent iterations. However, this model has not been applied to large-scale clusters. Cloud computing infrastructure offers computing resources as a collection of virtual machines by different hardware configurations, which is transparent to end users. In fact, the computational rates of these virtual machines instances are different and the system behaves as a heterogeneous environment. Thus, scheduling and load balancing for high performance computations become challenging issues.
We propose a hierarchical distributed approach suitable for scheduling parallel loops . We implemented our scheme on a large scale homogeneous cluster and also on a cloud system heterogeneous environment using the message passing interface protocol. We evaluated various performance aspects associated with our distributed scheduling scheme. In the future we plan to further investigate how cloud system features affect the performance of our approach and also to pursue implementation using the map reduce approach.