Scaling lattice QCD beyond 100 GPUs

Over the past five years, graphics processing units (GPUs) have had a transformational effect on numerical lattice quantum chromodynamics (LQCD) calculations in nuclear and particle physics. While GPUs have been applied with great success to the post-Monte Carlo "analysis" phase which acco...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2011 International Conference for High Performance Computing, Networking, Storage and Analysis (SC) s. 1 - 11
Hlavní autoři: Babich, R., Clark, M. A., Joó, B., Shi, G., Brower, R. C., Gottlieb, S.
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: New York, NY, USA ACM 12.11.2011
IEEE
Edice:ACM Conferences
Témata:
Applied computing > Physical sciences and engineering > Physics
Color
Computing methodologies > Symbolic and algebraic manipulation > Symbolic and algebraic algorithms > Linear algebra algorithms
domain decomposition
Equations
GPU
Graphics processing unit
Krylov solvers
Lattice QCD
Lattices
Libraries
Mathematical model
Mathematics of computing > Mathematical analysis > Differential equations > Partial differential equations
Mathematics of computing > Mathematical analysis > Numerical analysis > Computations in finite fields
Mathematics of computing > Mathematical analysis > Numerical analysis > Computations on matrices
Physics
Krylov solvers
GPU
domain decomposition
lattice QCD
ISBN:145030771X, 9781450307710
ISSN:2167-4329
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Over the past five years, graphics processing units (GPUs) have had a transformational effect on numerical lattice quantum chromodynamics (LQCD) calculations in nuclear and particle physics. While GPUs have been applied with great success to the post-Monte Carlo "analysis" phase which accounts for a substantial fraction of the workload in a typical LQCD calculation, the initial Monte Carlo "gauge field generation" phase requires capability-level supercomputing, corresponding to O(100) GPUs or more. Such strong scaling has not been previously achieved. In this contribution, we demonstrate that using a multi-dimensional parallelization strategy and a domain-decomposed preconditioner allows us to scale into this regime. We present results for two popular discretizations of the Dirac operator, Wilson-clover and improved staggered, employing up to 256 GPUs on the Edge cluster at Lawrence Livermore National Laboratory.
ISBN:145030771X
9781450307710
ISSN:2167-4329
DOI:10.1145/2063384.2063478