Filters








127 Hits in 3.3 sec

Load balancing of molecular dynamics simulation with NWChem

T. P. Straatsma, J. A. McCammon
2001 IBM Systems Journal  
This contribution provides details of the classical molecular dynamics module and focuses on issues related to load balancing on massively parallel computers, in particular the IBM SP TM and the Cray T3E  ...  This software integrates a range of modules for computational chemistry applications, including classical molecular dynamics simulations and quantum mechanical calculations.  ...  To make molecular simulations of such systems efficient, dynamic load-balancing techniques are needed.  ... 
doi:10.1147/sj.402.0328 fatcat:45of457fjvhmzpnsiftdliabia

Inspector-Executor Load Balancing Algorithms for Block-Sparse Tensor Contractions

David Ozog, Jeff R. Hammond, James Dinan, Pavan Balaji, Sameer Shende, Allen Malony
2013 2013 42nd International Conference on Parallel Processing  
Developing effective yet scalable load-balancing methods for irregular computations is critical to the successful application of simulations in a variety of disciplines at petascale and beyond.  ...  These two schemes, which have not been previously implemented in NWChem or its proxies (i.e. quantum chemistry mini-apps) are compared to the original centralized dynamic load-balancing algorithm as well  ...  NWChem NWChem [9] is the DOE flagship computational chemistry package, which supports most of the widely used methods across a range of accuracy scales (classical molecular dynamics, ab initio molecular  ... 
doi:10.1109/icpp.2013.12 dblp:conf/icpp/OzogHDBSM13 fatcat:an2u6ksugzgphdrhu6fcdgpooq

Inspector/executor load balancing algorithms for block-sparse tensor contractions

David Ozog, Sameer Shende, Allen Malony, Jeff R. Hammond, James Dinan, Pavan Balaji
2013 Proceedings of the 27th international ACM conference on International conference on supercomputing - ICS '13  
to centralized dynamic load balancing.  ...  Good load-balancing methods are required in order to obtain scalability from the NWChem coupled-cluster module, which allows the detailed study of chemical problems by iteratively solving the Schrödinger  ...  Scalability of centralized load-balancing The scalability of centralized dynamic load balancing with NXTVAL in the context of CC tensor contractions in NWChem CCSD simulation (15 iterations) with the aug-cc-pVDZ  ... 
doi:10.1145/2464996.2467282 dblp:conf/ics/OzogSMHDB13 fatcat:3jk6oyhbqngnzntnwescslxrb4

NWChem: New Functionality [chapter]

Theresa L. Windus, Eric J. Bylaska, Michel Dupuis, So Hirata, Lisa Pollack, Dayle M. Smith, T. P. Straatsma, Edoardo Aprà
2003 Lecture Notes in Computer Science  
Through effective use of the aggregate resources of such computers, NWChem enables much larger and more accurate production simulations than previously feasible.  ...  In this paper we briefly describe the NWChem architecture and highlight some of the new capabilities in NWChem.  ...  Molecular Dynamics Simulations of Biological Membranes The molecular dynamics (MD) simulation of biological membranes is generally more difficult to setup compared to simulations of other biological systems  ... 
doi:10.1007/3-540-44864-0_18 fatcat:ek3rogatfndahbinnfdhtqn2wq

NWChem: Past, Present, and Future [article]

E. Aprà and E. J. Bylaska and W. A. de Jong and N. Govind and K. Kowalski and T. P. Straatsma and M. Valiev and H. J. J. van Dam and Y. Alexeev and J. Anchell and V. Anisimov and F. W. Aquino and R. Atta-Fynn and J. Autschbach and N. P. Bauman and J. C. Becca and D. E. Bernholdt and K. Bhaskaran-Nair and S. Bogatko and P. Borowski and J. Boschen and J. Brabec and A. Bruner and E. Cauët and Y. Chen and G. N. Chuev and C. J. Cramer and J. Daily and M. J. O. Deegan and T. H. Dunning Jr. and M. Dupuis and K. G. Dyall and G. I. Fann and S. A. Fischer and A. Fonari and H. Früuchtl and L. Gagliardi and J. Garza and N. Gawande and S. Ghosh and K. Glaesemann and A. W. Götz and J. Hammond and V. Helms and E. D. Hermes and K. Hirao and S. Hirata and M. Jacquelin and L. Jensen and B. G. Johnson and H. Jónsson and R. A. Kendall and M. Klemm and R. Kobayashi and V. Konkov and S. Krishnamoorthy and M. Krishnan and Z. Lin and R. D. Lins and R. J. Littlefield and A. J. Logsdail and K. Lopata and W. Ma and A. V. Marenich and J. Martin del Campo and D. Mejia-Rodriguez and J. E. Moore and J. M. Mullin and T. Nakajima and D. R. Nascimento and J. A. Nichols and P. J. Nichols and J. Nieplocha and A. Otero de la Roza and B. Palmer and A. Panyala and T. Pirojsirikul and B. Peng and R. Peverati and J. Pittner and L. Pollack and R. M. Richard and P. Sadayappan and G. C. Schatz and W.A. Shelton and D. W. Silverstein and D. M. A. Smith and T. A. Soares and D. Song and M. Swart and H. L. Taylor and G. S. Thomas and V. Tipparaju and D. G. Truhlar and K. Tsemekhman and T. Van Voorhis and Á. Vázquez-Mayagoitia and P. Verma and O. Villa and A. Vishnu and K. D. Vogiatzis and D. Wang and J. H. Weare and M. J. Williamson and T. L. Windus and K. Woliński and A. T. Wong and Q. Wu and C. Yang and Q. Yu and M. Zacharias and Z. Zhang and Y. Zhao and R. J. Harrison
2020 Journal of Chemical Physics   accepted
and predictive many-body techniques that describe correlated behavior of electrons in molecular and condensed phase systems at different levels of theory.  ...  In enabling these simulations, novel parallel algorithms have been able to take advantage of computational resources to address the polynomial scaling of electronic structure methods.  ...  ACKNOWLEDGMENTS The authors acknowledge the important contributions to the NWChem project of the following deceased researchers: Ricky A. Kendall, Jarek Nieplocha, and Daniel W. Silverstein.  ... 
doi:10.1063/5.0004997 pmid:32414274 arXiv:2004.12023v2 fatcat:zzz2vczvkjbnjmqz2ssvd6ouku

NAMD: Biomolecular Simulation on Thousands of Processors

J.C. Phillips, Gengbin Zheng, S. Kumar, L.V. Kale
2002 ACM/IEEE SC 2002 Conference (SC'02)  
NAMD is a parallel, object-oriented molecular dynamics program designed for high performance simulation of large biomolecular systems.  ...  This paper discusses the techniques which have allowed NAMD to effectively employ over one thousand processors in production simulations of biomedical relevance.  ...  Importance of molecular dynamics simulation software.  ... 
doi:10.1109/sc.2002.10019 dblp:conf/sc/PhillipsZKK02 fatcat:slik5zdeqvbzpm6h4722seroku

A New Scalable Parallel Algorithm for Fock Matrix Construction

Xing Liu, Aftab Patel, Edmond Chow
2014 2014 IEEE 28th International Parallel and Distributed Processing Symposium  
Fine-grained tasks are better for load balance, but coarse-grained tasks require less communication.  ...  Existing algorithms and software, however, may fail to scale for large numbers of cores of a distributed machine, particularly in the simulation of moderately-sized molecules.  ...  A value of 1.000 indicates perfect load balance.  ... 
doi:10.1109/ipdps.2014.97 dblp:conf/ipps/LiuPC14 fatcat:zwhkhcj7hnbohmc3evmcjvneem

Performance Analysis of the NWChem TCE for Different Communication Patterns [chapter]

Priyanka Ghosh, Jeff R. Hammond, Sayan Ghosh, Barbara Chapman
2014 Lecture Notes in Computer Science  
We show that apart from the classical approach involving dynamic scheduling of data-blocks for load-balancing, reordering one-sided Get and Accumulate calls affects the performance of tensor contractions  ...  GA is used in a number of application codes, especially NWChem, and provides a superset of SHMEM functionality that includes remote accumulate, among other features.  ...  This research used resources of the Argonne Leadership Computing Facility at Argonne National Laboratory, which is supported by the Office of Science of the U.S.  ... 
doi:10.1007/978-3-319-10214-6_14 fatcat:b2cknz7vv5ednlta6a3pm2tbnm

WorkQ: A many-core producer/consumer execution model applied to PGAS computations

David Ozog, Allen Malony, Jeff R. Hammond, Pavan Balaji
2014 2014 20th IEEE International Conference on Parallel and Distributed Systems (ICPADS)  
Load balance, synchronization, and overlap of communication and computation are facilitated by a tunable nodewise FIFO message queue protocol.  ...  (step 3); but processes can switch roles dynamically for the sake of performance.  ...  NWChem is known for its strong capabilities in ab initio quantum chemistry methods invoking molecular electron structure theory, yet it also provides modules for simulating classical molecular dynamics  ... 
doi:10.1109/padsw.2014.7097863 dblp:conf/icpads/OzogMHB14 fatcat:dhtsfj4ozbb7nm62adnctznxly

Hard scaling challenges forab initiomolecular dynamics capabilities in NWChem: Using 100,000 CPUs per second

Eric J Bylaska, Kevin Glass, Doug Baxter, Scott B Baden, John H Weare
2009 Journal of Physics, Conference Series  
An overview of the parallel algorithms for ab initio molecular dynamics (AIMD) used in the NWChem program package is presented, including recent developments for computing exact exchange.  ...  Hard scaling challenges for ab initio molecular dynamics capabilities in NWChem: Using 100,000 CPUs per second To cite this article: Eric J Bylaska et al 2009 J. Phys.: Conf.  ...  The results shown are from NWChem ab initio molecular dynamics simulation of UO 2 2+ and 122 H 2 O molecules [2] .  ... 
doi:10.1088/1742-6596/180/1/012028 fatcat:xvts22vttjck3izu5uxocgnaue

Exploiting variability for energy optimization of parallel programs

Wim Lavrijsen, Costin Iancu, Wibe de Jong, Xin Chen, Karsten Schwan
2016 Proceedings of the Eleventh European Conference on Computer Systems - EuroSys '16  
By recognizing the "shape of variability" we can optimize codes with highly dynamic behavior, which pose challenges to all existing DVFS techniques.  ...  With NWChem on MPI two-sided and offline analysis, capturing the initialization, we find energy savings of up to 20%, with less than 1% performance cost. Publication rights licensed to ACM.  ...  Dynamic Behavior of NWChem: Slack is virtually nonexistent, as a result of dynamic load balancing in NWChem.  ... 
doi:10.1145/2901318.2901329 dblp:conf/eurosys/LavrijsenIJCS16 fatcat:kh3p5fzmqff3pkqelrjisuw724

High performance computations using dynamical nucleation theory

T L Windus, S M Kathmann, L D Crosby
2008 Journal of Physics, Conference Series  
Chemists continue to explore the use of very large computations to perform simulations that describe the molecular level physics of critical challenges in science.  ...  Disciplines Chemistry Comments This article is from Abstract: Chemists continue to explore the use of very large computations to perform simulations that describe the molecular level physics of critical  ...  It tolerates or hides latency by virtualizing the threads, facilitating the migration of the threads and provides dynamic load balancing to distribute work across the algorithm.  ... 
doi:10.1088/1742-6596/125/1/012017 fatcat:akhhbjl2qfccjjppu5sbfqoxci

Performance Tuning of Fock Matrix and Two-Electron Integral Calculations for NWChem on Leading HPC Platforms [chapter]

Hongzhang Shan, Brian Austin, Wibe De Jong, Leonid Oliker, N. J. Wright, Edoardo Apra
2014 Lecture Notes in Computer Science  
Our optimization schemes leverage key architectural features including vectorization and simultaneous multithreading, and results in speedups of up to 2.5x compared with the original implementation.  ...  Attaining performance in the evaluation of twoelectron repulsion integrals and constructing the Fock matrix is of considerable importance to the computational chemistry community.  ...  Task Granularity for Dynamic Load Balancing Profiling results indicate that the non-perfect linear scaling behavior is related with the dynamic load balancing approach.  ... 
doi:10.1007/978-3-319-10214-6_13 fatcat:wkirbsbqlbel5fylapjbbxiowy

A Hartree-Fock Application Using UPC++ and the New DArray Library

David Ozog, Amir Kamil, Yili Zheng, Paul Hargrove, Jeff R. Hammond, Allen Malony, Wibe de Jong, Kathy Yelick
2016 2016 IEEE International Parallel and Distributed Processing Symposium (IPDPS)  
The GTFock code is an HF implementation that, while it does not have all the features in NWChem, represents crucial algorithmic advances that reduce communication and improve load balance by doing an up-front  ...  We analyze the performance of atomic fetch-and-add operations (relevant for load balancing) and runtime attentiveness, then compare various techniques and optimizations for each.  ...  NWChem balances load with a centralized dynamic scheduler, which is controlled by a single global counter control referred to as nxtval (for "next value").  ... 
doi:10.1109/ipdps.2016.108 dblp:conf/ipps/OzogKZHHMJY16 fatcat:2kj32jz7grextjtzctsrbcoe6m

Barrier elision for production parallel programs

Milind Chabbi, Wim Lavrijsen, Wibe de Jong, Koushik Sen, John Mellor-Crummey, Costin Iancu
2015 SIGPLAN notices  
Overall, we demonstrate the value of holistic context-sensitive analyses that consider the domain science in conjunction with the associated runtime software stack.  ...  In this paper, we present context-sensitive dynamic optimizations that elide barriers redundant during the program execution.  ...  Department of Energy, Office of Science, Advanced Scientific Computing Research under collaborative agreement numbers DE-SC0008699 and DE-SC0010473.  ... 
doi:10.1145/2858788.2688502 fatcat:ijdxemwqcvfefihilzvif3hyn4
« Previous Showing results 1 — 15 out of 127 results