arrow
返回

A massively parallel tensor contraction framework for coupled-cluster computations

delete2014-12-01
delete148
PRE
AI
E
Edgar Solomonik *
D
Devin A. Matthews
J
Jeff R. Hammond
J
John F. Stanton
J
James Demmel
DOI:10.1016/j.jpdc.2014.06.002delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
Precise calculation of molecular electronic wavefunctions by methods such as coupled-cluster requires the computation of tensor contractions, the cost of which has polynomial computational scaling with respect to the system and basis set sizes. Each contraction may be executed via matrix multiplication on a properly ordered and structured tensor. However, data transpositions are often needed to reorder the tensors for each contraction. Writing and optimizing distributed-memory kernels for each transposition and contraction is tedious since the number of contractions scales combinatorially with the number of tensor indices. We present a distributed-memory numerical library (Cyclops Tensor Framework (CTF)) that automatically manages tensor blocking and redistribution to perform any user-specified contractions. CTF serves as the distributed-memory contraction engine in Aquarius, a new program designed for high-accuracy and massively-parallel quantum chemical computations. Aquarius implements a range of coupled-cluster and related methods such as CCSD and CCSDT by writing the equations on top of a C++ templated domain-specific language. This DSL calls CTF directly to manage the data and perform the contractions. Our CCSD and CCSDT implementations achieve high parallel scalability on the BlueGene/Q and Cray XC30 supercomputer architectures showing that accurate electronic structure calculations can be effectively carried out on top of general distributed-memory tensor primitives. (C) 2014 Elsevier Inc. All rights reserved.
Keyword:
Coupled-cluster
Tensor contractions
Matrix multiplication
Topology-aware mapping
Communication-avoiding algorithms

期刊

Journal of Parallel and Distributed Computing 封面图
Journal of Parallel and Distributed Computing
IF:
4
论文数:
3.8K
被引数:
4.8K

机构

U
University of California Berkeley
学者数:
3.5W
论文数: 2.8W
被引数: 11.3W
U
university of texas system
学者数:
18.5W
论文数: 15.6W
被引数: 210
University of California System 封面图
University of California System
学者数:
37.5W
论文数: 33.7W
被引数: 6.6K
学者 查看更多机构
引用论文

引用论文

err分享
err收藏
Synthesis of high-performance parallel programs for a class of Ab Initio quantum chemistry models一类从头算量子化学模型的高性能并行程序的合成
err2005-02-01
err139
PREAI
errBaumgartner, G; Auer, AA; Bernholdt, DE; Bibireata, A; Choppella, V; Cociorva, D; Gao, XY; Harrison, RJ; Hirata, S; Krishnamoorthy, S; Krishnan, S; Lam, CC; Lu, QD; Nooijen, M; Pitzer, RM; Ramanujam, J; Sadayappan, P; Sibiryakov, A
err分享
err收藏
err1996-01-01
err0
errOAAI
errLeo Breiman
err分享
err收藏
Divergent Semantic Behavior in Aphasia失语症的不同语义行为
err1976-12-01
err0
PREAI
errRoberta Chapey; Seymour Rigrodsky; Eleanor B. Morrison
err分享
err收藏
In vitro evaluation of anticancer and antibacterial activities of cobalt oxide nanoparticles
err2015-11-12
err0
PREAI
errShahanavaj Khan; Anees A. Ansari; Abdul Arif Khan; Rehan Ahmad; Omar Al-Obaid; Wael Al-Kattan
err分享
err收藏
学者 查看更多内容