Acyclic partitioning of large directed acyclic graphs

Herrmann, Julien and Kho, Jonathan and Uçar, Bora and Kaya, Kamer and Çatalyürek, Ümit (2017) Acyclic partitioning of large directed acyclic graphs. In: 17th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID 2017), Madrid, Spain

[thumbnail of 07973723.pdf] PDF
07973723.pdf
Restricted to Repository staff only

Download (383kB) | Request a copy

Abstract

Finding a good partition of a computational directed acyclic graph associated with an algorithm can help find an execution pattern improving data locality, conduct an analysis of data movement, and expose parallel steps. The partition is required to be acyclic, i.e., the inter-part edges between the vertices from different parts should preserve an acyclic dependency structure among the parts. In this work, we adopt the multilevel approach with coarsening, initial partitioning, and refinement phases for acyclic partitioning of directed acyclic graphs and develop a direct k-way partitioning scheme. To the best of our knowledge, no such scheme exists in the literature. To ensure the acyclicity of the partition at all times, we propose novel and efficient coarsening and refinement heuristics. The quality of the computed acyclic partitions is assessed by computing the edge cut, the total volume of communication between the parts, and the critical path latencies. We use the solution returned by well-known undirected graph partitioners as a baseline to evaluate our acyclic partitioner, knowing that the space of solution is more restricted in our problem. The experiments are run on large graphs arising from linear algebra applications.
Item Type: Papers in Conference Proceedings
Subjects: Q Science > QA Mathematics > QA075 Electronic computers. Computer science
Divisions: Faculty of Engineering and Natural Sciences > Academic programs > Computer Science & Eng.
Faculty of Engineering and Natural Sciences
Depositing User: Kamer Kaya
Date Deposited: 14 Sep 2017 23:00
Last Modified: 26 Apr 2022 09:27
URI: https://research.sabanciuniv.edu/id/eprint/32990

Actions (login required)

View Item
View Item