Search

Your search keyword '"Communication Time"' showing total 127 results
Start Over Descriptor "Communication Time"

Refine your results

more »

more »

more »
127 results on '"Communication Time"'

1. Finding broadcast 2-centers of a tree under the postal model.

Author

Tsou, Cheng-Hsiao, Lin, Ching-Chi, and Hsu, Chan-Hung

Subjects
*BROADCASTING studios, *BROADCASTING industry, *PROBLEM solving, *TREES
Abstract

In this paper, we delve into the investigation of locating broadcast 2-centers of a tree T under the postal model. The problem asks to deploy two broadcast centers so that the maximum communication time from the centers to their corresponding vertices is minimized. The broadcast 2-center problem explored in this study is essentially an extension of the classical 2-center problem. To efficiently address this challenge, we employ the binary search method, which enables us to present a highly effective O (n log n) -time algorithm for solving this problem. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

2. A scheduling policy to save 10% of communication time in parallel fast Fourier transform.

Author

Aseeri, Samar A., Gopal Chatterjee, Anando, Verma, Mahendra K., and Keyes, David E.

Subjects
FAST Fourier transforms, PARTIAL differential equations, DENSITY functional theory, RADIO astronomy, FLUID flow, SIGNAL processing
Abstract

Summary: The fast Fourier transform (FFT) has applications in almost every frequency related study, for example, in image and signal processing, and radio astronomy. It is also used as a Poisson operator inversion kernel in partial differential equations in fluid flows, in density functional theory, many‐body theory, and others. The three‐dimensional N3 FFT has large time complexity O(N3log2N). Hence, parallelization is used to compute such FFTs. Popular libraries perform slab division or pencil decomposition of N3 data. None of the existing libraries achieve perfect inverse scaling of time with (T−1≈n) cores because FFT requires all‐to‐all communication and clusters hitherto do not have physical all‐to‐all connections. Dragonfly, one of the popular topologies for the interconnect, supports hierarchical connections among the components. We show that if we align the all‐to‐all communication of FFT with the physical connections of Dragonfly topology we will achieve a better scaling and reduce communication time. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

3. Computation And Communication Efficient Chinese Remainder Theorem Based Multi-Party Key Generation Using Modified RSA

Author

Rawat, Arjun Singh, Deshmukh, Maroti, Angrisani, Leopoldo, Series Editor, Arteaga, Marco, Series Editor, Panigrahi, Bijaya Ketan, Series Editor, Chakraborty, Samarjit, Series Editor, Chen, Jiming, Series Editor, Chen, Shanben, Series Editor, Chen, Tan Kay, Series Editor, Dillmann, Rüdiger, Series Editor, Duan, Haibin, Series Editor, Ferrari, Gianluigi, Series Editor, Ferre, Manuel, Series Editor, Hirche, Sandra, Series Editor, Jabbari, Faryar, Series Editor, Jia, Limin, Series Editor, Kacprzyk, Janusz, Series Editor, Khamis, Alaa, Series Editor, Kroeger, Torsten, Series Editor, Liang, Qilian, Series Editor, Martín, Ferran, Series Editor, Ming, Tan Cher, Series Editor, Minker, Wolfgang, Series Editor, Misra, Pradeep, Series Editor, Möller, Sebastian, Series Editor, Mukhopadhyay, Subhas, Series Editor, Ning, Cun-Zheng, Series Editor, Nishida, Toyoaki, Series Editor, Pascucci, Federica, Series Editor, Qin, Yong, Series Editor, Seng, Gan Woon, Series Editor, Speidel, Joachim, Series Editor, Veiga, Germano, Series Editor, Wu, Haitao, Series Editor, Zhang, Junjie James, Series Editor, Stănică, Pantelimon, editor, Gangopadhyay, Sugata, editor, and Debnath, Sumit Kumar, editor

Published
2021
Full Text
View/download PDF

4. The Role of Communication Time in the Convergence of Federated Edge Learning.

Author

Zhou, Yipeng, Fu, Yao, Luo, Zhenxiao, Hu, Miao, Wu, Di, Sheng, Quan Z., and Yu, Shui

Subjects
*REACTION time, *SMARTPHONES, *STIMULUS & response (Psychology), *EDGES (Geometry), *COST functions, *MACHINE learning
Abstract

Federated Edge Learning (FEL) enables a massive number of edge devices (e.g. smart phones) to train machine learning models collaboratively. Due to the inherent unreliability of participating edge devices and the unpredictable network conditions, the final model accuracy is largely determined by the communication time between edge devices and the parameter server (PS). However, there exists very limited work to quantify the influence of the communication time in FEL and show how to minimize its negative impacts from a theoretic perspective. In this paper, we are among the first to develop a formal model of the communication time in FEL and its influence on the final model accuracy. In our work, the set of edge devices involved in each global iteration is defined as the ECP (Engaged Client Pool). We model the communication time cost as a function with respect to the response time distribution of individual devices and the ECP size. By incorporating communication time cost into the convergence rate analysis, we propose the ECPA and H-ECPA algorithms to automatically adjust the size of the ECP so as to maximize the model accuracy in both homogeneous and heterogeneous networks. We also analyze how the tail shape of response time affects the convergence rate, and prove that the heavy tail shape can significantly lower the model accuracy. Finally, we conduct extensive experiments with real datasets, and the results confirm the correctness of our analysis and demonstrate the superiority of our proposed algorithms. [ABSTRACT FROM AUTHOR]

Published
2022
Full Text
View/download PDF

5. Decentralised federated learning with adaptive partial gradient aggregation

Author

Jingyan Jiang and Liang Hu

Subjects
learning (artificial intelligence), gradient methods, communication frequency, parameter server design, nodes-to-server bandwidths, stochastic gradient descent training, end-to-end training, real-world federated learning scenarios, adaptive partial gradient aggregation method, gradient partial level decentralised, partial gradient exchange mechanism, node-to-node bandwidth, communication time, adaptive model, training time, decentralised federated learning, machine learning model, geo-distributed workers, inherently communication, communication efficiency, conventional federated learning algorithms, Computational linguistics. Natural language processing, P98-98.5, Computer software, QA76.75-76.765
Abstract

Federated learning aims to collaboratively train a machine learning model with possibly geo-distributed workers, which is inherently communication constrained. To achieve communication efficiency, the conventional federated learning algorithms allow the worker to decrease the communication frequency by training the model locally for multiple times. Conventional federated learning architecture, inherited from the parameter server design, relies on highly centralised topologies and large nodes-to-server bandwidths, and convergence property relies on the stochastic gradient descent training in local, which usually causes the large end-to-end training latency in real-world federated learning scenarios. Thus, in this study, the authors propose the adaptive partial gradient aggregation method, a gradient partial level decentralised federated learning, to tackle this problem. In FedPGA, they propose a partial gradient exchange mechanism that makes full use of node-to-node bandwidth for speeding up the communication time. Besides, an adaptive model updating method further reduces the convergence rate by adaptive increasing the step size of the stable direction of gradient descent. The experimental results on various datasets demonstrate that the training time is reduced up to [inline-formula] compared to baselines without accuracy degrade.

Published
2020
Full Text
View/download PDF

6. Decentralised federated learning with adaptive partial gradient aggregation.

Author

Jiang, Jingyan and Hu, Liang

Abstract

Federated learning aims to collaboratively train a machine learning model with possibly geo‐distributed workers, which is inherently communication constrained. To achieve communication efficiency, the conventional federated learning algorithms allow the worker to decrease the communication frequency by training the model locally for multiple times. Conventional federated learning architecture, inherited from the parameter server design, relies on highly centralised topologies and large nodes‐to‐server bandwidths, and convergence property relies on the stochastic gradient descent training in local, which usually causes the large end‐to‐end training latency in real‐world federated learning scenarios. Thus, in this study, the authors propose the adaptive partial gradient aggregation method, a gradient partial level decentralised federated learning, to tackle this problem. In FedPGA, they propose a partial gradient exchange mechanism that makes full use of node‐to‐node bandwidth for speeding up the communication time. Besides, an adaptive model updating method further reduces the convergence rate by adaptive increasing the step size of the stable direction of gradient descent. The experimental results on various datasets demonstrate that the training time is reduced up to 14× compared to baselines without accuracy degrade. [ABSTRACT FROM AUTHOR]

Published
2020
Full Text
View/download PDF

9. A Map-Reduce Parallel Approach to Automatic Synthesis of Control Software

Author

Alimguzhin, Vadim, Mari, Federico, Melatti, Igor, Salvo, Ivano, Tronci, Enrico, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Doug, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Bartocci, Ezio, editor, and Ramakrishnan, C. R., editor

Published
2013
Full Text
View/download PDF

10. Optimisation of Patch Distribution Strategies for AMR Applications

Author

Beckingsale, D. A., Perks, O. F. J., Gaudin, W. P., Herdman, J. A., Jarvis, S. A., Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Doug, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Tribastone, Mirco, editor, and Gilmore, Stephen, editor

Published
2013
Full Text
View/download PDF

11. Micro-applications for Communication Data Access Patterns and MPI Datatypes

Author

Schneider, Timo, Gerstenberger, Robert, Hoefler, Torsten, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Doug, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Träff, Jesper Larsson, editor, Benkner, Siegfried, editor, and Dongarra, Jack J., editor

Published
2012
Full Text
View/download PDF

12. Hybrid Parallelization of a Large-Scale Heart Model

Author

Krause, Dorian, Potse, Mark, Dickopf, Thomas, Krause, Rolf, Auricchio, Angelo, Prinzen, Frits, Hutchison, David, Series editor, Kanade, Takeo, Series editor, Kittler, Josef, Series editor, Kleinberg, Jon M., Series editor, Mattern, Friedemann, Series editor, Mitchell, John C., Series editor, Naor, Moni, Series editor, Nierstrasz, Oscar, Series editor, Pandu Rangan, C., Series editor, Steffen, Bernhard, Series editor, Sudan, Madhu, Series editor, Terzopoulos, Demetri, Series editor, Tygar, Doug, Series editor, Vardi, Moshe Y., Series editor, Weikum, Gerhard, Series editor, Keller, Rainer, editor, Kramer, David, editor, and Weiss, Jan-Philipp, editor

Published
2012
Full Text
View/download PDF

15. A New Algorithm for Divisible Load Scheduling with Different Processor Available Times

Author

Shokripour, Amin, Othman, Mohamed, Ibrahim, Hamidah, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Doug, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Goebel, Randy, editor, Siekmann, Jörg, editor, Wahlster, Wolfgang, editor, Nguyen, Ngoc Thanh, editor, Le, Manh Thanh, editor, and Świątek, Jerzy, editor

Published
2010
Full Text
View/download PDF

16. Data Transmission Scheme Considering Node Failure for Blockchain.

Author

Li, Jiao

Subjects
BLOCKCHAINS, DISTRIBUTED computing, DATA security, COMPUTER security, COMPUTER algorithms, COMPUTER system failures
Abstract

In recent years, Blockchain technology has attracted considerable attention from the industrial circle. Blockchain is a distributed ledger technology, which must be validated, stored and maintained by all nodes to ensure data security, transparency, and integrity. The communication efficiency of Blockchain is an important factor restricting its application. Existing algorithms can provide data routing schemes for Blockchain but without considering the node failure. On Blockchain, node failure is a common phenomenon due to the nodes’ selfishness and nodes’ mobility. Node failure degrades the network performance or even sometimes makes the network useless. This paper proposes a data transmission scheme considering node failure for finishing validation of block data on Blockchain, which firstly sets response threshold level to detect failure node, and then using greedy idea constructs communication tree to organize all nodes forwarding block data. Based to the multi-link concurrent communication tree model, this scheme maximizes the potential transmitting capacity of nodes and assigns proper tasks to other nodes beside source node, so it can shorten the validation time of Blockchain transaction, and improve resistance to node failure. Theoretical proof and experimental results show the effectiveness and the efficiency of the proposed data transmission scheme. [ABSTRACT FROM AUTHOR]

Published
2018
Full Text
View/download PDF

17. Increasing the Scalability and the Speedup of a Fish School Simulator

Author

Dalforno, Christianne, Mostaccio, Diego, Suppi, Remo, Luque, Emilio, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Doug, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Gervasi, Osvaldo, editor, Murgante, Beniamino, editor, Laganà, Antonio, editor, Taniar, David, editor, Mun, Youngsong, editor, and Gavrilova, Marina L., editor

Published
2008
Full Text
View/download PDF

18. Multiple Access Communications for Age Minimization in UAV Aided Data Collection

Author

Ogutcu O., Yuksel M., Ogutcu O., and Yuksel M.

Abstract

Huawei;Nokia;pix moving;Samsung;Technology Innovation Institute (TII), 95th IEEE Vehicular Technology Conference - Spring, VTC 2022-Spring -- 19 June 2022 through 22 June 2022 -- -- 182307, Data collection with unmanned aerial vehicles (UAVs) is a novel technology with many practical applications. In this paper, UAV aided data collection in sensor networks for time-critical applications is considered. Maximum age of information (AoI) is used as the performance metric, where both communication time and flight time contribute to AoI. Given that sensors form groups of two, the UAV aims to visit each line segment formed by the two sensors in each group once and to return to the data center. To minimize the maximum AoI, two trajectory design algorithms are proposed. In both algorithms, the UAV communicates with the sensors in each group through a multiple access channel. In the first algorithm, the UAV hovers on the communication time optimal point on the line segment. In the second algorithm, it hovers on the path-length optimal point. The algorithms are compared with time sharing schemes, and also with the case when all sensors are visited individually. The results reveal that multiple access transmission is optimal, when sensors are not too far apart and packet lengths are long, whereas path-length optimal trajectory is better when packet lengths are short. © 2022 IEEE.

Published
2022

19. Multiple Access Communications for Age Minimization in UAV Aided Data Collection

Author

Yuksel M., Ogutcu O., Yuksel M., and Ogutcu O.

Abstract

Huawei;Nokia;pix moving;Samsung;Technology Innovation Institute (TII), 95th IEEE Vehicular Technology Conference - Spring, VTC 2022-Spring -- 19 June 2022 through 22 June 2022 -- -- 182307, Data collection with unmanned aerial vehicles (UAVs) is a novel technology with many practical applications. In this paper, UAV aided data collection in sensor networks for time-critical applications is considered. Maximum age of information (AoI) is used as the performance metric, where both communication time and flight time contribute to AoI. Given that sensors form groups of two, the UAV aims to visit each line segment formed by the two sensors in each group once and to return to the data center. To minimize the maximum AoI, two trajectory design algorithms are proposed. In both algorithms, the UAV communicates with the sensors in each group through a multiple access channel. In the first algorithm, the UAV hovers on the communication time optimal point on the line segment. In the second algorithm, it hovers on the path-length optimal point. The algorithms are compared with time sharing schemes, and also with the case when all sensors are visited individually. The results reveal that multiple access transmission is optimal, when sensors are not too far apart and packet lengths are long, whereas path-length optimal trajectory is better when packet lengths are short. © 2022 IEEE.

Published
2022

20. Parallel Dynamic SPT Update Algorithm in OSPF

Author

Zhu, Yuanbo, Xu, Mingwei, Wu, Qian, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Doug, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, and Malyshkin, Victor, editor

Published
2007
Full Text
View/download PDF

21. Optimizing Evolution Rules Application and Communication Times in Membrane Systems Implementation

Author

Tejedor, Jorge A., Gutiérrez, Abraham, Fernández, Luis, Arroyo, Fernando, Bravo, Ginés, Gómez, Sandra, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Doug, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Eleftherakis, George, editor, Kefalas, Petros, editor, Păun, Gheorghe, editor, Rozenberg, Grzegorz, editor, and Salomaa, Arto, editor

Published
2007
Full Text
View/download PDF

22. Parallelization Strategies

Author

Chattot, J. -J., editor, Colella, P., editor, Weinan, E., editor, Glowinski, R., editor, Holt, M., editor, Hussaini, Y., editor, Joly, P., editor, Keller, H. B., editor, Marsden, J. E., editor, Meiron, D. I., editor, Pironneau, O., editor, Quarteroni, A., editor, Rappaz, J., editor, Rosner, R., editor, Sagaut, P., editor, Seinfeld, J. H., editor, Szepessy, A., editor, Wheeler, M. F., editor, Schneider, Johannes Josef, and Kirkpatrick, Scott

Published
2006
Full Text
View/download PDF

23. Modeling Message-Passing Overhead on NCHC Formosa PC Cluster

Author

Chou, Chau-Yi, Chang, Hsi-Ya, Wang, Shuen-Tai, Tcheng, Shou-Cheng, Hutchison, David, Series editor, Kanade, Takeo, Series editor, Kittler, Josef, Series editor, Kleinberg, Jon M., Series editor, Mattern, Friedemann, Series editor, Mitchell, John C., Series editor, Naor, Moni, Series editor, Nierstrasz, Oscar, Series editor, Pandu Rangan, C., Series editor, Steffen, Bernhard, Series editor, Sudan, Madhu, Series editor, Terzopoulos, Demetri, Series editor, Tygar, Dough, Series editor, Vardi, Moshe Y., Series editor, Weikum, Gerhard, Series editor, Chung, Yeh-Ching, editor, and Moreira, José E., editor

Published
2006
Full Text
View/download PDF

24. Web-Based Cooperative Design for SoC and Improved Architecture Exploration Algorithm

Author

Ma, Guangsheng, Wang, Xiuqin, Wang, Hao, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Dough, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Shen, Heng Tao, editor, Li, Jinbao, editor, Li, Minglu, editor, Ni, Jun, editor, and Wang, Wei, editor

Published
2006
Full Text
View/download PDF

25. Analytical Models of Probability Distributions for MPI Point-to-Point Communication Times on Distributed Memory Parallel Computers

Author

Grove, D. A., Coddington, P. D., Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Dough, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Hobbs, Michael, editor, Goscinski, Andrzej M., editor, and Zhou, Wanlei, editor

Published
2005
Full Text
View/download PDF

26. Automatic Tuning of Master/Worker Applications

Author

Morajko, Anna, César, Eduardo, Caymes-Scutari, Paola, Margalef, Tomás, Sorribes, Joan, Luque, Emilio, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Dough, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Cunha, José C., editor, and Medeiros, Pedro D., editor

Published
2005
Full Text
View/download PDF

27. A Performance Prediction Model for Tomographic Reconstruction in Structural Biology

Author

Fritzsche, Paula Cecilia, Fernández, José-Jesús, Ripoll, Ana, García, Inmaculada, Luque, Emilio, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Dough, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Daydé, Michel, editor, Dongarra, Jack, editor, Hernández, Vicente, editor, and Palma, José M. L. M., editor

Published
2005
Full Text
View/download PDF

28. High Frequency Electromagnetic Fields Analysis with the Use of the Parallel FDTD Method

Author

Walendziuk, Wojciech, Forenc, Jaroslaw, Jordan, Andrzej, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Terzopoulos, Demetri, editor, Tygar, Dough, editor, Vardi, Moshe Y., editor, Wyrzykowski, Roman, editor, Dongarra, Jack, editor, Paprzycki, Marcin, editor, and Waśniewski, Jerzy, editor

Published
2004
Full Text
View/download PDF

36. A Columnwise Block Striping in Neville Elimination

Author

Alonso, Pedro, Cortina, Raquel, Díaz, Irene, Hernández, Vicente, Ranilla, José, Goos, Gerhard, editor, Hartmanis, Juris, editor, van Leeuwen, Jan, editor, Wyrzykowski, Roman, editor, Dongarra, Jack, editor, Paprzycki, Marcin, editor, and Waśniewski, Jerzy, editor

Published
2002
Full Text
View/download PDF

45. Predictive communication modeling for HPC applications.

Author

Papadopoulou, Nikela, Goumas, Georgios, and Koziris, Nectarios

Subjects
*COMMUNICATION models, *PREDICTION models, *COMMUNICATION, *MACHINE learning, *COMPUTER systems, *SUPERCOMPUTERS
Abstract

In this paper, we present a methodology for predictive modeling of communication of HPC applications. Communication time depends on a complex set of parameters, relevant to the application, the system architecture, the runtime configuration and runtime conditions. To handle this complexity, we define features that can be extracted from the application, the process mapping and the allocation shape ahead of execution, deploy a single benchmark to sweep over the parameter space and develop predictive models for communication time on two supercomputers, Vilje and Piz Daint, using different subsets of our features, machine-learning methods and training sets. We compare the predictive power of our models on two common communication patterns and one application, for various problem sizes, executions and runtime configurations, ranging from a few dozen to a few thousand cores. Our methodology is successful across all tested communication patterns on both systems and exhibits high prediction accuracy and goodness-of-fit, scoring 23.98% in MMRE, 0.942 in RCC and 61.43% in $$Pred_{0.25}$$ on Vilje and 21.31%, 0.940 and 66.57% respectively on Piz Daint, with models that are applicable just-in-time ahead of the execution of an HPC application. [ABSTRACT FROM AUTHOR]

Published
2017
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results