We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Distributed, Parallel, and Cluster Computing

Title: Parallel Graph Coloring Algorithms for Distributed GPU Environments

Abstract: Graph coloring is often used in parallelizing scientific computations that run in distributed and multi-GPU environments; it identifies sets of independent data that can be updated in parallel. Many algorithms exist for graph coloring on a single GPU or in distributed memory, but to the best of our knowledge, hybrid MPI+GPU algorithms have been unexplored until this work. We present several MPI+GPU coloring approaches based on the distributed coloring algorithms of Gebremedhin et al. and the shared-memory algorithms of Deveci et al. . The on-node parallel coloring uses implementations in KokkosKernels,
which provide parallelization for both multicore CPUs and GPUs. We further extend our approaches to compute distance-2 and partial distance-2 colorings, giving the first known distributed, multi-GPU algorithm for these problems. In addition, we propose a novel heuristic to reduce communication for recoloring in distributed graph coloring. Our experiments show that our approaches operate efficiently on inputs too large to fit on a single GPU and scale up to graphs with 76.7 billion edges running on 128 GPUs.
Comments: Submitted to Parallel Computing
Subjects: Distributed, Parallel, and Cluster Computing (cs.DC); Discrete Mathematics (cs.DM)
Cite as: arXiv:2107.00075 [cs.DC]
  (or arXiv:2107.00075v1 [cs.DC] for this version)

Submission history

From: Erik Boman [view email]
[v1] Wed, 30 Jun 2021 19:55:57 GMT (1289kb,D)

Link back to: arXiv, form interface, contact.