Готові списки джерел за темами / Distribution shared memory

Добірка наукової літератури з теми "Distribution shared memory"

Автор: Grafiati

Опубліковано: 14 березня 2023

Оформте джерело за APA, MLA, Chicago, Harvard та іншими стилями

Оберіть тип джерела:

Зміст

Статті в журналах
Дисертації
Частини книг
Тези доповідей конференцій
Звіти організацій

Ознайомтеся зі списками актуальних статей, книг, дисертацій, тез та інших наукових джерел на тему "Distribution shared memory".

Біля кожної праці в переліку літератури доступна кнопка «Додати до бібліографії». Скористайтеся нею – і ми автоматично оформимо бібліографічне посилання на обрану працю в потрібному вам стилі цитування: APA, MLA, «Гарвард», «Чикаго», «Ванкувер» тощо.

Також ви можете завантажити повний текст наукової публікації у форматі «.pdf» та прочитати онлайн анотацію до роботи, якщо відповідні параметри наявні в метаданих.

Статті в журналах з теми "Distribution shared memory"

Chandra, Rohit, Ding-Kai Chen, Robert Cox, Dror E. Maydan, Nenad Nedeljkovic, and Jennifer M. Anderson. "Data distribution support on distributed shared memory multiprocessors." ACM SIGPLAN Notices 32, no. 5 (May 1997): 334–45. http://dx.doi.org/10.1145/258916.258945.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Marzolla, Moreno, and Gabriele D’angelo. "Parallel Data Distribution Management on Shared-memory Multiprocessors." ACM Transactions on Modeling and Computer Simulation 30, no. 1 (February 8, 2020): 1–25. http://dx.doi.org/10.1145/3369759.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Axtmann, Michael, Sascha Witt, Daniel Ferizovic, and Peter Sanders. "Engineering In-place (Shared-memory) Sorting Algorithms." ACM Transactions on Parallel Computing 9, no. 1 (March 31, 2022): 1–62. http://dx.doi.org/10.1145/3505286.

Повний текст джерела

Анотація:

We present new sequential and parallel sorting algorithms that now represent the fastest known techniques for a wide range of input sizes, input distributions, data types, and machines. Somewhat surprisingly, part of the speed advantage is due to the additional feature of the algorithms to work in-place, i.e., they do not need a significant amount of space beyond the input array. Previously, the in-place feature often implied performance penalties. Our main algorithmic contribution is a blockwise approach to in-place data distribution that is provably cache-efficient. We also parallelize this approach taking dynamic load balancing and memory locality into account. Our new comparison-based algorithm In-place Parallel Super Scalar Samplesort ( IPS 4 o ) , combines this technique with branchless decision trees. By taking cases with many equal elements into account and by adapting the distribution degree dynamically, we obtain a highly robust algorithm that outperforms the best previous in-place parallel comparison-based sorting algorithms by almost a factor of three. That algorithm also outperforms the best comparison-based competitors regardless of whether we consider in-place or not in-place, parallel or sequential settings. Another surprising result is that IPS 4 o even outperforms the best (in-place or not in-place) integer sorting algorithms in a wide range of situations. In many of the remaining cases (often involving near-uniform input distributions, small keys, or a sequential setting), our new In-place Parallel Super Scalar Radix Sort ( IPS 2 Ra ) turns out to be the best algorithm. Claims to have the – in some sense – “best” sorting algorithm can be found in many papers which cannot all be true. Therefore, we base our conclusions on an extensive experimental study involving a large part of the cross product of 21 state-of-the-art sorting codes, 6 data types, 10 input distributions, 4 machines, 4 memory allocation strategies, and input sizes varying over 7 orders of magnitude. This confirms the claims made about the robust performance of our algorithms while revealing major performance problems in many competitors outside the concrete set of measurements reported in the associated publications. This is particularly true for integer sorting algorithms giving one reason to prefer comparison-based algorithms for robust general-purpose sorting.

Стилі APA, Harvard, Vancouver, ISO та ін.

O'Hearn, Kurt A., Abdullah Alperen, and Hasan Metin Aktulga. "Fast Solvers for Charge Distribution Models on Shared Memory Platforms." SIAM Journal on Scientific Computing 42, no. 1 (January 2020): C1—C22. http://dx.doi.org/10.1137/18m1224684.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

BYNA, SURENDRA, KIRK W. CAMERON, and XIAN-HE SUN. "ISOLATING COSTS IN SHARED MEMORY COMMUNICATION BUFFERING." Parallel Processing Letters 15, no. 04 (December 2005): 357–65. http://dx.doi.org/10.1142/s0129626405002271.

Повний текст джерела

Анотація:

Communication in parallel applications is a combination of data transfers internally at a source or destination and across the network. Previous research focused on quantifying network transfer costs has indirectly resulted in reduced overall communication cost. Optimized data transfer from source memory to the network interface has received less attention. In shared memory systems, such memory-to-memory transfers dominate communication cost. In distributed memory systems, memory-to-network interface transfers grow in significance as processor and network speeds increase at faster rates than memory latency speeds. Our objective is to minimize the cost of internal data transfers. The following examples illustrating the impact of memory transfers on communication, we present a methodology for classifying the effects of data size and data distribution on hardware, middleware, and application software performance. This cost is quantified using hardware counter event measurements on the SGI Origin 2000. For the SGI O2K, we empirically identify the cost caused by just copying data from one buffer to another and the middleware overhead. We use MPICH in our experiments, but our techniques are generally applicable to any communication implementation.

Стилі APA, Harvard, Vancouver, ISO та ін.

Addison, C., Y. Ren, and M. van Waveren. "OpenMP Issues Arising in the Development of Parallel BLAS and LAPACK Libraries." Scientific Programming 11, no. 2 (2003): 95–104. http://dx.doi.org/10.1155/2003/278167.

Повний текст джерела

Анотація:

Dense linear algebra libraries need to cope efficiently with a range of input problem sizes and shapes. Inherently this means that parallel implementations have to exploit parallelism wherever it is present. While OpenMP allows relatively fine grain parallelism to be exploited in a shared memory environment it currently lacks features to make it easy to partition computation over multiple array indices or to overlap sequential and parallel computations. The inherent flexible nature of shared memory paradigms such as OpenMP poses other difficulties when it becomes necessary to optimise performance across successive parallel library calls. Notions borrowed from distributed memory paradigms, such as explicit data distributions help address some of these problems, but the focus on data rather than work distribution appears misplaced in an SMP context.

Стилі APA, Harvard, Vancouver, ISO та ін.

Crooks, P., and R. H. Perrott. "Language Constructs for Data Partitioning and Distribution." Scientific Programming 4, no. 2 (1995): 59–85. http://dx.doi.org/10.1155/1995/656010.

Повний текст джерела

Анотація:

This article presents a survey of language features for distributed memory multiprocessor systems (DMMs), in particular, systems that provide features for data partitioning and distribution. In these systems the programmer is freed from consideration of the low-level details of the target architecture in that there is no need to program explicit processes or specify interprocess communication. Programs are written according to the shared memory programming paradigm but the programmer is required to specify, by means of directives, additional syntax or interactive methods, how the data of the program are decomposed and distributed.

Стилі APA, Harvard, Vancouver, ISO та ін.

Nikolopoulos, Dimitrios S., Ernest Artiaga, Eduard Ayguadé, and Jesús Labarta. "Scaling Non-Regular Shared-Memory Codes by Reusing Custom Loop Schedules." Scientific Programming 11, no. 2 (2003): 143–58. http://dx.doi.org/10.1155/2003/379739.

Повний текст джерела

Анотація:

In this paper we explore the idea of customizing and reusing loop schedules to improve the scalability of non-regular numerical codes in shared-memory architectures with non-uniform memory access latency. The main objective is to implicitly setup affinity links between threads and data, by devising loop schedules that achieve balanced work distribution within irregular data spaces and reusing them as much as possible along the execution of the program for better memory access locality. This transformation provides a great deal of flexibility in optimizing locality, without compromising the simplicity of the shared-memory programming paradigm. In particular, the programmer does not need to explicitly distribute data between processors. The paper presents practical examples from real applications and experiments showing the efficiency of the approach.

Стилі APA, Harvard, Vancouver, ISO та ін.

Bozkus, Zeki, Larry Meadows, Steven Nakamoto, Vincent Schuster, and Mark Young. "PGHPF – An Optimizing High Performance Fortran Compiler for Distributed Memory Machines." Scientific Programming 6, no. 1 (1997): 29–40. http://dx.doi.org/10.1155/1997/705102.

Повний текст джерела

Анотація:

High Performance Fortran (HPF) is the first widely supported, efficient, and portable parallel programming language for shared and distributed memory systems. HPF is realized through a set of directive-based extensions to Fortran 90. It enables application developers and Fortran end-users to write compact, portable, and efficient software that will compile and execute on workstations, shared memory servers, clusters, traditional supercomputers, or massively parallel processors. This article describes a production-quality HPF compiler for a set of parallel machines. Compilation techniques such as data and computation distribution, communication generation, run-time support, and optimization issues are elaborated as the basis for an HPF compiler implementation on distributed memory machines. The performance of this compiler on benchmark programs demonstrates that high efficiency can be achieved executing HPF code on parallel architectures.

Стилі APA, Harvard, Vancouver, ISO та ін.

Warren, Karen H. "PDDP, A Data Parallel Programming Model." Scientific Programming 5, no. 4 (1996): 319–27. http://dx.doi.org/10.1155/1996/857815.

Повний текст джерела

Анотація:

PDDP, the parallel data distribution preprocessor, is a data parallel programming model for distributed memory parallel computers. PDDP implements high-performance Fortran-compatible data distribution directives and parallelism expressed by the use of Fortran 90 array syntax, the FORALL statement, and the WHERE construct. Distributed data objects belong to a global name space; other data objects are treated as local and replicated on each processor. PDDP allows the user to program in a shared memory style and generates codes that are portable to a variety of parallel machines. For interprocessor communication, PDDP uses the fastest communication primitives on each platform.

Стилі APA, Harvard, Vancouver, ISO та ін.

Більше джерел

Дисертації з теми "Distribution shared memory"

Leventhal, Sean. "Speculative data distribution in shared memory multiprocessors." College Park, Md. : University of Maryland, 2008. http://hdl.handle.net/1903/8076.

Повний текст джерела

Анотація:

Thesis (Ph. D.) -- University of Maryland, College Park, 2008.
Thesis research directed by: Dept. of Electrical and Computer Engineering. Title from t.p. of PDF. Includes bibliographical references. Published by UMI Dissertation Services, Ann Arbor, Mich. Also available in paper.

Стилі APA, Harvard, Vancouver, ISO та ін.

HEMMATPOUR, MASOUD. "High Performance Computing using Infiniband-based clusters." Doctoral thesis, Politecnico di Torino, 2019. http://hdl.handle.net/11583/2750549.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Hedberg, Per Henrik. "Interpersonal society : essays on shared beliefs, trust, mnemonic oppression, distributive fairness, and value creation." Doctoral thesis, Handelshögskolan i Stockholm, Institutionen för Marknadsföring och strategi, 2012. http://urn.kb.se/resolve?urn=urn:nbn:se:hhs:diva-1761.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Trapanese, Cinzia. "Spatial foraging in primates : strategies and mechanisms of decision-making What, where and when: spatial foraging decisions in primates Species-specific socio-ecology shapes spatial foraging strategies in primates Where or what? Primates in “miniature nature”: frugivory triggers spatial cognition to forage efficiently." Thesis, Sorbonne Paris Cité, 2018. http://www.theses.fr/2018USPCB119.

Повний текст джерела

Анотація:

Chercher sa nourriture s'avère coûteux et potentiellement exigeant cognitivement, notamment pour des animaux comme les primates, vivant dans un environnement difficilement prévisible. Les fruits étant une ressource éphémère, un régime alimentaire frugivore implique des capacités cognitives élevées. Cette étude comparative s'intéresse aux différentes stratégies mises en place lors du fourragement chez trois espèces de primates (Macaca tonkeana=5, M. fascicularis=3 and Sapajus sp.=6) vivant dans des parcs boisés, au Centre de Primatologie de l'Université de Strasbourg. Ces conditions d'hébergement en semi-liberté nous ont permis de manipuler l'environnement afin de simuler la répartition saisonnière spatio-temporelle des fruits en milieu naturel. Nous avons notamment fait varier la disponibilité de différents fruits chaque semaine avec un rythme saisonnier d'un mois, répété pendant quatre mois. Notre protocole expérimental vise à évaluer le poids des variables Où (localisation, tâche « Fourragement Spatial »), Quoi (distribution, tâches « Groupé vs. Dispersé » et qualité, tâche « Groupé vs. Qualité »), et Quand (disponibilité) sur les décisions individuelles lors de la recherche de nourriture. Nous avons utilisé 42 boîtes à ouverture télécommandée, fixées sur des arbres de l'enclos des singes, dans lesquelles nous avons placé des fruits. Ce dispositif a permis de tester les sujets individuellement dans leur groupe social. Pour chaque essai, nous avons enregistré les trajectoires des sujets, l'ordre des boîtes visitées et la présence de congénères. Dans la tâche « Fourragement Spatial », nous avons comparé les trajectoires observées des essais dans lesquels les sujets ont visité toutes les boites (Ntot=77) avec des trajectoires simulées pour trois stratégies : optimale, de proche en proche, ou aléatoire. Aucune espèce n'a suivi une stratégie aléatoire, les capucins suivent fréquemment une trajectoire optimale, les macaques fascicularis suivent davantage une stratégie de proche en proche alors que les Tonkeans suivent les deux stratégies avec des fréquences similaires. Toutes les espèces ont adopté soit une stratégie globale (trajet optimal) soit une stratégie locale (boite la plus proche) afin de fourrager efficacement. Les expériences de la tâche « Groupé vs. Dispersé » proposent aux animaux de choisir parmi des boîtes dont la répartition spatiale forme deux patchs de nourriture de six boites chacun, l'un groupé, l'autre dispersé. Nous avons testé l'effet de cette distribution spatiale en utilisant les mêmes fruits dans les deux patchs (Nobs=2477). Nos résultats indiquent une préférence pour la distribution groupée (vs. dispersée) chez les trois espèces. Nous avons également testé l'effet de la qualité de la nourriture (tâche « Groupé vs. Qualité ») en proposant deux fruits différents disponibles en même temps, le préféré dans la distribution dispersée vs. le moins préféré dans la distribution groupée (Nobs=2546). Les trois espèces ont continué de choisir préférentiellement la distribution groupée, mais les Tonkeans dans une moindre mesure que les capucins, ce qui indique que la qualité influence également leurs choix. Les Tonkeans ont aussi montré des déplacements plus directs. Enfin, pour la composante Quand, les sujets ont bien adapté leur fourragement à la séquence temporelle de la disponibilité alimentaire, répétée quatre fois, mais des hypothèses alternatives peuvent expliquer ces résultats (e.g. capacité à reconnaitre les boîtes disponibles). Cette étude montre que toutes les espèces utilisent des stratégies de fourragement efficaces. Cependant, contrairement aux omnivores, les frugivores ont davantage pris en compte leurs préférences alimentaires et ont montré des déplacements plus directs vers les sites alimentaires. Cette étude souligne que les contraintes écologiques des espèces peuvent affecter l'évolution des capacités cognitives et, plus généralement, des comportements
Foraging can be a challenging activity, especially for animals like primates living in seasonal environments characterised by not fairly predictable food availability. Since fruit is an ephemeral resource, a frugivorous diet is associated to brain size and high cognitive abilities. This comparative study aims to investigate the spatial foraging strategies of three primate species (Macaca tonkeana=5 individuals, M. fascicularis=3 and Sapajus sp.=6) having different degree of frugivory and living in semi-free ranging conditions at the Primate Centre of Strasbourg University. The experimental protocol aims to assess the relative weight of Where (food location, Spatial Foraging Task), Where vs. What (food distribution, Clumped vs. Scattered Tasks and quality, Clumped vs. Quality Task) and When (food temporal availability) variables on the individual foraging decisions. Forty-two boxes were fixed on trees in the outdoor area and were lockable via a remote-control system to individually test subjects in their social group. Each week, a subset of boxes was filled with one/two appealing fruit types; a seasonal pace of one month was repeated for four months to mimic the seasonality of wild fruit. We recorded subjects' trajectories, the order of visited boxes and the presence of other individuals. In the Spatial Foraging Task, considering the trials in which subjects visited all six baited boxes (Tonkeans Ntrials=35, long-tailed macaques Ntrials=31, capuchins Ntrials=11), we compared the observed routes to simulated routes under three strategies: optimal route, nearest neighbour rule, random route. None of the species choose random routes, suggesting that they relied on spatial memory to visit food sites. Capuchins optimized more their travels than long-tailed macaques, which followed mainly a nearest neighbour strategy, while Tonkeans used both strategies at similar frequency. All study species used a global (optimal path) or local strategy (nearest-neighbour rule) to forage efficiently. In the Clumped vs. Scattered Tasks, we tested how the food distribution influences primate spatial foraging. In the Clumped vs. Scattered Task 1, 12 boxes were baited with the same fruit type, six boxes in a circular clumped distribution and other six in a scattered circle. The Task 2 provided a similarly preferred fruit in the same two circular configurations but with opposite reciprocal spatial positions in terms of the side of the outdoor area. All study species (Nvisited boxes=2477) visited at first significantly more the clumped distribution. In the Clumped vs. Quality Task, we assessed the relative impact of food preference versus food distribution: boxes had the similar spatial configuration of the Clumped vs. Scattered Task 1 but the scattered distribution was filled with the most favourite fruit and the clumped one with the least preferred fruit. All species (Nvisited boxes=2546) showed again a preference for visiting first the boxes of the clumped distribution, but the most frugivorous species, Tonkeans, showed a less strong preference compared to the least frugivorous, the capuchins. The higher was the frugivory degree of the species, the higher were the goal-directed travels. Lastly, to investigate if primates developed a temporal knowledge of fruit availability, we investigated if they correctly remembered food spatio-temporal availability: primates visited each month between 79%-98% of baited boxes/tot boxes visited, even if this positive result could be explained by alternative hypothesis (e.g. ability in detecting the available boxes). All primate species maximised foraging efficiency, avoiding random walks. However, frugivorous species took food preference into account in their decisions and showed significantly more goal-directed movement. This study underlines how species feeding ecology may affect the evolution of their abstract mental abilities and more in general, their behaviour

Стилі APA, Harvard, Vancouver, ISO та ін.

Huang, Chi-Ting, and 黃啟庭. "A Resource-Oriented Workload Distribution Scheme for Software Distributed Shared Memory Systems." Thesis, 2003. http://ndltd.ncl.edu.tw/handle/99763665405419622002.

Повний текст джерела

Анотація:

碩士
國立成功大學
電機工程學系碩博士班
91
Dynamic workload distribution plays an important role in Software Distributed Shared Memory Systems to improve system performance efficiently. It can adjust the workload distribution of DSM applications with runtime collected information at the executing period. By adequately utilizing various resources among nodes, DSM applications can archive the load balance and get the best performance. In early researches, most of them can not process workload distribution well among nodes which have different resource capability. It is because that the factors affecting the performance are not only the processor power but also the available physical memory and network bandwidth. By taking all of these factors into consideration, the DSM system can provide better performance in most situations. In this thesis, we will introduce a workload distribution scheme, named Resource-Oriented Workload Distribution Scheme (ROWDS), to achieve high performance in different environments. In the experiments, we can see that our ROWDS, which considers several factors simultaneously, can perform better than the other ones merely taking a single factor into account.

Стилі APA, Harvard, Vancouver, ISO та ін.

Kok, Yih-Tzye, and 郭義潔. "Design and Implementation of Load Distributing on Distributed Shared Memory System." Thesis, 1995. http://ndltd.ncl.edu.tw/handle/88777216904949625158.

Повний текст джерела

Анотація:

碩士
國立成功大學
電機工程研究所
83
Synchronization barries are the most costly constructs among other synchronization constructs in the MIMD parallel program model. The idle time due to synchronization barrier can be a source of overhead in a parallel computation, particularly when some load unbalance occurs. In this thesis, we will examine the effects of load unbalance on the synchronization barriers, and attempt to proposed a dynamic load distributing algorithm to circumvent these unbalance effects by minimizes the idle times that incurred by synchronization barriers. Additionally, communication cost due to the dependencies between threads are important consideration issues for performance improvement in the design of DSM system. We have attempted to include the communication cost consideration into our load distributing system design, by introduce a selection policy a selection policy for selecting a suitable thread for migration, which will reduce the communication cost in the system. All of the algorithms have been implemented on a locally distributed shared memory system, Cohesion. The experimental results show that the effect of load unbalance to the synchronization barrier will be reduced significantly, while our selection policy only do well on certain applications with a regular shared data access pattern.

Стилі APA, Harvard, Vancouver, ISO та ін.

Частини книг з теми "Distribution shared memory"

Nikolopoulos, Dimitrios S., and Eduard Ayguadé. "A Study of Implicit Data Distribution Methods for OpenMP Using the SPEC Benchmarks." In OpenMP Shared Memory Parallel Programming, 115–29. Berlin, Heidelberg: Springer Berlin Heidelberg, 2001. http://dx.doi.org/10.1007/3-540-44587-0_11.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Ayguadé, Eduard, Jordi Garcia, M. Luz Grande, and Jesús Labarta. "Data distribution and loop parallelization for shared-memory multiprocessors." In Languages and Compilers for Parallel Computing, 41–55. Berlin, Heidelberg: Springer Berlin Heidelberg, 1997. http://dx.doi.org/10.1007/bfb0017244.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Hirooka, Takashi, Hiroshi Ohta, and Takayoshi Iitsuka. "Automatic Data Distribution Method Using First Touch Control for Distributed Shared Memory Multiprocessors." In Languages and Compilers for Parallel Computing, 147–61. Berlin, Heidelberg: Springer Berlin Heidelberg, 2003. http://dx.doi.org/10.1007/3-540-35767-x_10.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Mattingly, R. B., and C. D. Meyer. "Computing the Stationary Distribution Vector of an Irreducible Markov Chain on a Shared-Memory Multiprocessor." In Numerical Solution of Markov Chains, 491–510. Boca Raton: CRC Press, 2021. http://dx.doi.org/10.1201/9781003210160-26.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Shanmugan, S. "Influences of Powder Size (SMAs) Distribution Fe–Mn/625 Alloy Systematic Studies of 4D-Printing Conceivable Applications." In Shape Memory Composites Based on Polymers and Metals for 4D Printing, 81–92. Cham: Springer International Publishing, 2022. http://dx.doi.org/10.1007/978-3-030-94114-7_5.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Dudai, Yadin. "Persistence of Collective Memory over 3,000 Years." In National Memories, 259–79. Oxford University PressNew York, 2022. http://dx.doi.org/10.1093/oso/9780197568675.003.0013.

Повний текст джерела

Анотація:

Abstract Collective memory is a set of historical narratives, beliefs, and customs shared by a social group, such as a community, culture, or nation, over generations. This chapter presents observations concerning the collective memory of the Jewish culture from the vantage point of the science of memory. Evidence for what later came to be regarded as Jewish culture can be traced back more than 3,200 years (i.e., more than 130 generations) ago. The early history of the culture amalgamated fact with fiction over scores of generations in orally reliant communities before being put in writing more than 2,300 years ago in a textual epitome, or credo, of only 63 Hebrew words. The long-term cultural persistence of this foundation core of the collective memory was set at the outset to rely on procedures to ensure regular semantic recitation combined with episodic re-enactment. Since then and up to the present time, memories of a number of major collective traumas have been added to the repertoire of Jewish collective memory. In recent centuries, the ancient credo has contributed to the revitalization and realization of a national movement; yet in doing so, it has also contributed to a rather fast evolution of Jewish collective memory, manifested in its ongoing differentiation into subnarratives that differ, inter alia, in their attitudes toward nationalism and in geographical distribution, religious hue, and populist flavor.

Стилі APA, Harvard, Vancouver, ISO та ін.

Bisseling, Rob H. "Sparse matrix–vector multiplication." In Parallel Scientific Computation, 190–290. Oxford University Press, 2020. http://dx.doi.org/10.1093/oso/9780198788348.003.0004.

Повний текст джерела

Анотація:

This chapter introduces irregular algorithms and presents the example of parallel sparse matrix-vector multiplication (SpMV), which is the central operation in iterative linear system solvers. The irregular sparsity pattern of the matrix does not change during the multiplication, which may be repeated many times. This justifies putting a lot of effort into finding a good data distribution. The Mondriaan distribution of a sparse matrix is a useful non-Cartesian distribution that can be found by hypergraph-based partitioning. The Mondriaan package implements such a partitioning and also the newer medium-grain partitioning method. The chapter analyses the special cases of random sparse matrices and Laplacian matrices. It uses performance profiles and geometric means to compare different partitioning methods. Furthermore, it presents the hybrid-BSP model and a hybrid-BSP SpMV, which are aimed at hybrid distributed/shared-memory architectures. The parallel SpMV can be incorporated in applications, ranging from PageRank computation to artificial neural networks.

Стилі APA, Harvard, Vancouver, ISO та ін.

Yang, Ning, Shiaaulir Wang, and Paul Schonfeld. "Simulation-Based Scheduling of Waterway Projects Using a Parallel Genetic Algorithm." In Transportation Systems and Engineering, 334–47. IGI Global, 2015. http://dx.doi.org/10.4018/978-1-4666-8473-7.ch016.

Повний текст джерела

Анотація:

A Parallel Genetic Algorithm (PGA) is used for a simulation-based optimization of waterway project schedules. This PGA is designed to distribute a Genetic Algorithm application over multiple processors in order to speed up the solution search procedure for a very large combinational problem. The proposed PGA is based on a global parallel model, which is also called a master-slave model. A Message-Passing Interface (MPI) is used in developing the parallel computing program. A case study is presented, whose results show how the adaption of a simulation-based optimization algorithm to parallel computing can greatly reduce computation time. Additional techniques which are found to further improve the PGA performance include: (1) choosing an appropriate task distribution method, (2) distributing simulation replications instead of different solutions, (3) avoiding the simulation of duplicate solutions, (4) avoiding running multiple simulations simultaneously in shared-memory processors, and (5) avoiding using multiple processors which belong to different clusters (physical sub-networks).

Стилі APA, Harvard, Vancouver, ISO та ін.

Yang, Ning, Shiaaulir Wang, and Paul Schonfeld. "Simulation-Based Scheduling of Waterway Projects Using a Parallel Genetic Algorithm." In Civil and Environmental Engineering, 1071–84. IGI Global, 2016. http://dx.doi.org/10.4018/978-1-4666-9619-8.ch046.

Повний текст джерела

Анотація:

Стилі APA, Harvard, Vancouver, ISO та ін.

Petersen, Wesley, and Peter Arbenz. "Shared Memory Parallelism." In Introduction to Parallel Computing. Oxford University Press, 2004. http://dx.doi.org/10.1093/oso/9780198515760.003.0009.

Повний текст джерела

Анотація:

Shared memory machines typically have relatively few processors, say 2–128. An intrinsic characteristic of these machines is a strategy for memory coherence and a fast tightly coupled network for distributing data from a commonly accessible memory system. Our test examples were run on two HP Superdome clusters: Stardust is a production machine with 64 PA-8700 processors, and Pegasus is a 32 CPU machine with the same kind of processors. The HP9000 is grouped into cells, each with 4 CPUs, a common memory/cell, and connected to a CCNUMA crossbar network. The network consists of sets of 4×4 crossbars and is shown in Figure 4.2. An effective bandwidth test, the EFF_BW benchmark [116], groups processors into two equally sized sets. Arbitrary pairings are made between elements from each group, Figure 4.3, and the cross-sectional bandwidth of the network is measured for a fixed number of processors and varying message sizes. The results from the HP9000 machine Stardust are shown in Figure 4.4. It is clear from this figure that the cross-sectional bandwidth of the network is quite high. Although not apparent from Figure 4.4, the latency for this test (the intercept near Message Size = 0) is not high. Due to the low incremental resolution of MPI_Wtime, multiple test runs must be done to quantify the latency. Dr Byrde’s tests show that minimum latency is ≳ 1.5μs. A clearer example of a shared memory architecture is the Cray X1 machine, shown in Figures 4.5 and 4.6. In Figure 4.6, the shared memory design is obvious. Each multi-streaming processor (MSP) shown in Figure 4.5 has 4 processors (custom designed processor chips forged by IBM), and 4 corresponding caches. Although not clear from available diagrams, vector memory access apparently permits cache by-pass; hence the term streaming in MSP. That is, vector registers are loaded directly from memory: see, for example, Figure 3.4. On each board (called nodes) are 4 such MSPs and 16 memory modules which share a common (coherent) memory view. Coherence is only maintained on each board, but not across multiple board systems.

Стилі APA, Harvard, Vancouver, ISO та ін.

Тези доповідей конференцій з теми "Distribution shared memory"

Chandra, Rohit, Ding-Kai Chen, Robert Cox, Dror E. Maydan, Nenad Nedeljkovic, and Jennifer M. Anderson. "Data distribution support on distributed shared memory multiprocessors." In the ACM SIGPLAN 1997 conference. New York, New York, USA: ACM Press, 1997. http://dx.doi.org/10.1145/258915.258945.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Nodine, Mark H., and Jeffrey Scott Vitter. "Deterministic distribution sort in shared and distributed memory multiprocessors." In the fifth annual ACM symposium. New York, New York, USA: ACM Press, 1993. http://dx.doi.org/10.1145/165231.165247.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Marzolla, Moreno, and Gabriele D'Angelo. "Parallel sort-based matching for data distribution management on shared-memory multiprocessors." In 2017 IEEE/ACM 21st International Symposium on Distributed Simulation and Real-Time Applications (DS-RT). IEEE, 2017. http://dx.doi.org/10.1109/distra.2017.8167660.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Wolin, Elliott, D. Abbott, V. Gurjyan, G. Heyes, E. Jastrzembski, D. Lawrence., and C. Timmer. "The ET system - high speed event transfer and distribution via shared memory and networks." In 2007 IEEE Nuclear Science Symposium Conference Record. IEEE, 2007. http://dx.doi.org/10.1109/nssmic.2007.4436464.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Nikolopoulos, Dimitrios S., Eduard Ayguadé, Theodore S. Papatheodorou, Constantine D. Polychronopoulos, and Jesús Labarta. "The trade-off between implicit and explicit data distribution in shared-memory programming paradigms." In the 15th international conference. New York, New York, USA: ACM Press, 2001. http://dx.doi.org/10.1145/377792.377801.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Dzafic, I., and H. T. Neisius. "Real-time power flow algorithm for shared memory multiprocessors for European distribution network types." In Energy Conference (IPEC 2010). IEEE, 2010. http://dx.doi.org/10.1109/ipecon.2010.5697097.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Chung, Soon M., and Jaerheen Yang. "Distributive join algorithm for shared-memory multiprocessors." In the 1993 ACM/SIGAPP symposium. New York, New York, USA: ACM Press, 1993. http://dx.doi.org/10.1145/162754.168683.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Duan, Shanzhong, and Andrew Ries. "Efficient Parallel Computer Simulation of the Motion Behaviors of Closed-Loop Multibody Systems." In ASME 2007 International Mechanical Engineering Congress and Exposition. ASMEDC, 2007. http://dx.doi.org/10.1115/imece2007-41912.

Повний текст джерела

Анотація:

This paper presents an efficient parallelizable algorithm for the computer-aided simulation and numerical analysis of motion behaviors of multibody systems with closed-loops. The method is based on cutting certain user-defined system interbody joints so that a system of independent multibody subchains is formed. These subchains interact with one another through associated unknown constraint forces fc at the cut joints. The increased parallelism is obtainable through cutting joints and the explicit determination of associated constraint forces combined with a sequential O(n) method. Consequently, the sequential O(n) procedure is carried out within each subchain to form and solve the equations of motion while parallel strategies are performed between the subchains to form and solve constraint equations concurrently. For multibody systems with closed-loops, joint separations play both a role of creation of parallelism for computing load distribution and a role of opening a closed-loop for use of the O(n) algorithm. Joint separation strategies provide the flexibility for use of the algorithm so that it can easily accommodate the available number of processors while maintaining high efficiency. The algorithm gives the best performance for the application scenarios for n>>1 and n>>m, where n and m are number of degree of freedom and number of constraints of a multibody system with closed-loops respectively. The algorithm can be applied to both distributed-memory parallel computing systems and shared-memory parallel computing systems.

Стилі APA, Harvard, Vancouver, ISO та ін.

Lowenthal, D. K., V. W. Freeh, and D. W. Miller. "Efficient support for two-dimensional data distributions in distributed shared memory systems." In Proceedings 16th International Parallel and Distributed Processing Symposium. IPDPS 2002. IEEE, 2002. http://dx.doi.org/10.1109/ipdps.2002.1015552.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Jovanova, Jovana, Angela Nastevska, and Mary Frecker. "Target Shape Optimization of 3D Compliant Mechanism With Superelastic Joints and Shape Memory Actuation." In ASME 2019 Conference on Smart Materials, Adaptive Structures and Intelligent Systems. American Society of Mechanical Engineers, 2019. http://dx.doi.org/10.1115/smasis2019-5639.

Повний текст джерела

Анотація:

Abstract The shape memory effect and the superelasticity of nickel titanium (NiTi) alloys are beneficial for design of compliant mechanisms. The superelastic behavior of NiTi can be tailored for optimal flexure design in the compliant mechanism, allowing large deformation and shape change. The shape memory effect can also be utilized to actuate the compliant mechanism flexures enabling programing of the material to take on variety of shapes at different temperatures over time. The compliant mechanism analyzed in this work is inspired from 3D multi leg spider-like locomotion, enabling movement in all directions by triggering different target shapes in time. The control of the material spatial distribution facilitated by additive manufacturing will enable tailored superelastic and shape memory behavior in the flexures of the multifunctional 3D compliant mechanism. Design optimization and analyses as well as overall shape change are explored in this work. Superelastic joints are introduced as flexures to enable segment flexibility. The temperature change is used for actuation taking in consideration different initial strain conditions.

Стилі APA, Harvard, Vancouver, ISO та ін.

Звіти організацій з теми "Distribution shared memory"

Merritt, Alexander M., and Kevin Thomas Tauke Pedretti. LDRD final report : managing shared memory data distribution in hybrid HPC applications. Office of Scientific and Technical Information (OSTI), September 2010. http://dx.doi.org/10.2172/1007320.

Повний текст джерела

Стилі APA, Harvard, Vancouver, ISO та ін.

Ми пропонуємо знижки на всі преміум-плани для авторів, чиї праці увійшли до тематичних добірок літератури. Зв'яжіться з нами, щоб отримати унікальний промокод!