Paper Info
Title
Accelerating lattice QCD multigrid on GPUs using fine-grained parallelization.
Abstract
The past decade has witnessed a dramatic acceleration of lattice quantum chromodynamics calculations in nuclear and particle physics. This has been due to both significant progress in accelerating the iterative linear solvers using multi-grid algorithms, and due to the throughput improvements brought by GPUs. Deploying hierarchical algorithms optimally on GPUs is non-trivial owing to the lack of parallelism on the coarse grids, and as such, these advances have not proved multiplicative. Using the QUDA library, we demonstrate that by exposing all sources of parallelism that the underlying stencil problem possesses, and through appropriate mapping of this parallelism to the GPU architecture, we can achieve high efficiency even for the coarsest of grids. Results are presented for the Wilson-Clover discretization, where we demonstrate up to 10x speedup over present state-of-the-art GPU-accelerated methods on Titan. Finally, we look to the future, and consider the software implications of our findings.
Year
DOI
Venue
2016
10.1109/SC.2016.67
SC
Keywords
Field
DocType
lattice QCD multigrid,fine-grained parallelization,lattice quantum chromodynamics calculations,nuclear physics,particle physics,iterative linear solvers,multigrid algorithms,Titan,Wilson-Clover discretization,GPU architecture,stencil problem,QUDA library,coarse grids,hierarchical algorithms
Discretization,Computer performance,CUDA,Computer science,Stencil,Parallel computing,Lattice QCD,PCI Express,Multigrid method,Speedup
Conference
ISSN
ISBN
Citations 
Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC '16), Article 68 (November, 2016)
978-1-4673-8815-3
0
PageRank 
References 
Authors
0.34
11
6
Name
Order
Citations
PageRank
Michael A. Clark1475.19
B. Joó2768.67
A. Strelchenko331.49
Michael Cheng400.34
Arjun Gambhir500.34
Richard C. Brower651.49