On the Scalability of an Automatically Parallelized Irregular Application

Martin Burtscher, Milind Kulkarni, Dimitrios Prountzos, and Keshav Pingali

LCPC 2008 · International Workshop on Languages and Compilers for Parallel Computing · August 2008 · pp. 109-123

Acceptance rate 51%

Abstract

Irregular applications, i.e., programs that manipulate pointer-based data structures such as graphs and trees, constitute a challenging target for parallelization because the amount of parallelism is input dependent and changes dynamically. Traditional dependence analysis techniques are too conservative to expose this parallelism. Even manual parallelization is difficult, time consuming, and error prone. The Galois system parallelizes such applications using an optimistic approach that exploits higher-level semantics of abstract data types.

In this paper, we study the performance and scalability of a Galoised, i.e., automatically parallelized, version of Delaunay mesh refinement (DR) on a shared-memory system with 128 CPUs. DR is an important irregular application that is used, e.g., in graphics and finite-element codes. The parallelized program scales to 64 threads, where it reaches a speedup of 25.8. For large numbers of threads, the performance is hampered by the load imbalance and the nonuniform memory latency, both of which grow as the number of threads increases. While these two issues will have to be addressed in future work, we believe our results already show the Galois approach to be very promising.

BibTeX

@inproceedings{burtscher08lcpc,
  author    = {Martin Burtscher and Milind Kulkarni and Dimitrios Prountzos and Keshav Pingali},
  title     = {On the Scalability of an Automatically Parallelized Irregular Application},
  booktitle = {Proceedings of the International Workshop on Languages and Compilers for Parallel Computing (LCPC)},
  pages     = {109--123},
  publisher = {Springer-Verlag},
  year      = {2008},
  month     = aug,
  doi       = {10.1007/978-3-540-89740-8_8}
}