Optimization of sparse matrix-vector multiplication on emerging multicore platforms
Open accessThe DOI matches its Crossref record; 21 of the 23 references it cites were found in a registry or an index.
Abstract
We are witnessing a dramatic change in computer architecture due to the multicore paradigm shift, as every electronic device from cell phones to supercomputers confronts parallelism of unprecedented scale. To fully unleash the potential of these systems, the HPC community must develop multicore specific optimization methodologies for important scientific computations. In this work, we examine sparse matrix-vector multiply (SpMV) - one of the most heavily used kernels in scientific computing - across a broad spectrum of multicore designs. Our experimental platform includes the homogeneous AMD dual-core and Intel quad-core designs, the heterogeneous STI Cell, as well as the first scientific study of the highly multithreaded Sun Niagara2. We present several optimization strategies especially effective for the multicore environment, and demonstrate significant performance improvements compared to existing state-of-the-art serial and parallel SpMV implementations. Additionally, we present key insights into the architectural tradeoffs of leading multicore design strategies, in the context of demanding memory-bound numerical algorithms.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Citations
Cites
Efficient Management of Parallelism in Object-Oriented Numerical Software Libraries
Balay, S., Gropp, W. D., McInnes, L. C., & Smith, B. F. (1997) · Modern Software Tools for Scientific Computing · doi:10.1007/978-1-4612-1986-6_8 · Cited by 2,009 · cited 15×
“When the matrix has dense block substructure, as is common .nite element method-based applications, practitioners employ a blocked variant of CSR (BCSR), which stores a single column index for each r × c block, rather than one per non-zero as in CSR. 2.1 OSKI, OSKI-PETSc, and RelatedWork Wecompareour multicoreSpMV op…”
BackgroundInfluentialDesign challenges of technology scaling
Borkar, S. (1999) · IEEE Micro · doi:10.1109/40.782564 · Cited by 1,164 · cited 1×
“Industry has moved to chip multiprocessor (CMP) system design in order to better manage trade-offs among performance, energy efficiency, and reliability [5, 9].”
BackgroundOptimization of sparse matrix–vector multiplication on emerging multicore platforms
Williams, S., Oliker, L., Vuduc, R., Shalf, J., Yelick, K., & Demmel, J. (2009) · Parallel Computing · doi:10.1016/j.parco.2008.12.006 · Cited by 789
Self-citationOSKI: A library of automatically tuned sparse matrix kernels
Vuduc, R., Demmel, J. W., & Yelick, K. A. (2005) · Journal of Physics: Conference Series · doi:10.1088/1742-6596/16/1/071 · Cited by 490 · cited 5×
“, symmetry, advanced register blocking, A methods [20]) as a key algorithmic optimization .”
BackgroundMethodResultInfluentialSelf-citationRose, D. J. (1972) · Elsevier eBooks · doi:10.1016/b978-1-4832-3187-7.50018-0 · Cited by 449 · cited 1×
“Besides classical bandwidth-reducing reordering techniques [15], recent work has proposed sophisticated 2-D partitioning schemes with theoretical guarantees on communication volume [18], and traveling salesman-based reordering to create dense block substructure [14].”
MethodSynergistic Processing in Cell's Multicore Architecture
Gschwind, M. K., Hofstee, H. P., Flachs, B., Hopkins, M. E., Watanabe, Y., & Yamazaki, T. (2006) · IEEE Micro · doi:10.1109/mm.2006.41 · Cited by 403 · cited 1×
“Cell adopts a heterogeneous approach to multicore, with one conventional processor core (Power Processing Element / PPE) to handle OS and control functions, combined with up to eight simpler SIMD cores (Synergistic Processing Elements / SPEs) for the computationally intensive work [7, 8].”
Sparsity: Optimization Framework for Sparse Matrix Kernels
Im, E.-J., Yelick, K., & Vuduc, R. (2004) · The International Journal of High Performance Computing Applications · doi:10.1177/1094342004041296 · Cited by 339 · cited 6×
“These patterns include blocks [10], variable or mixtures of differently-sized blocks [6] diagonals, which may be especially well-suited to machines with SIMD and vector units [19], dense subtriangles arising in sparse triangular solve [22], and symmetry [11], and combinations.”
BackgroundMethodInfluentialSelf-citationA Two-Dimensional Data Distribution Method for Parallel Sparse Matrix-Vector Multiplication
Vastenhouw, B., & Bisseling, R. H. (2005) · SIAM Review · doi:10.1137/s0036144502409019 · Cited by 274 · cited 1×
“Besides classical bandwidth-reducing reordering techniques [15], recent work has proposed sophisticated 2-D partitioning schemes with theoretical guarantees on communication volume [18], and traveling salesman-based reordering to create dense block substructure [14].”
MethodImproving performance of sparse matrix-vector multiplication
Pinar, A., & Heath, M. T. (1999) · Proceedings of the 1999 ACM/IEEE conference on Supercomputing · doi:10.1145/331532.331562 · Cited by 250
Improving the memory-system performance of sparse-matrix vector multiplication
Toledo, S. (1997) · IBM Journal of Research and Development · doi:10.1147/rd.416.0711 · Cited by 207 · cited 1×
“Recent work on low-level tuning of SpMV by unroll-and-jam [12], software pipelining [6], and prefetching [ 17 ] influence our work.”
Background
Research works printed but not linked: 7
- 10.5555/266469.266486 · doi:10.5555/266469.266486
- J. L. Hennessy and D. A. Patterson . Computer Architecture: A Quantitative Approach ; fourth edition. Morgan Kaufmann , San Francisco , 2006 . J. L. Hennessy and D. A. Patterson. Computer Architecture: A Quantitative Approach; fourth edition. Morgan Kaufmann, San Francisco, 2006.
- 10.5555/1018425.1020284 · doi:10.5555/1018425.1020284
- J. Mellor-Crummey and J. Garvin . Optimizing sparse matrix vector multiply using unroll-and-jam . In Proc. LACSI Symposium , Santa Fe, NM, USA , October 2002 . J. Mellor-Crummey and J. Garvin. Optimizing sparse matrix vector multiply using unroll-and-jam. In Proc. LACSI Symposium, Santa Fe, NM, USA, October 2002.
- R. Vuduc , A. Gyulassy , J. W. Demmel , and K. A. Yelick . Memory hierarchy optimizations and bounds for sparse ATAx . In Proceedings of the ICCS Workshop on Parallel Linear Algebra , volume LNCS, Melbourne, Australia, June 2003 . Springer. R. Vuduc, A. Gyulassy, J. W. Demmel, and K. A. Yelick. Memory hierarchy optimizations and bounds for sparse ATAx. In Proceedings of the ICCS Workshop on Parallel Linear Algebra, volume LNCS, Melbourne, Australia, June 2003. Springer.
- J. W. Willenbring , A. A. Anda , and M. Heroux . Improving sparse matrix-vector product kernel performance and availabillity . In Proc. Midwest Instruction and Computing Symposium, Mt . Pleasant, IA , 2006 . J. W. Willenbring, A. A. Anda, and M. Heroux. Improving sparse matrix-vector product kernel performance and availabillity. In Proc. Midwest Instruction and Computing Symposium, Mt. Pleasant, IA, 2006.
- 10.1017/s10766-007-0034-5 · doi:10.1017/s10766-007-0034-5
Cited by
Items 701–800 of 810
Toward compiler-driven adaptive execution and its application to GPU architectures
Lee, S. (2011) · Purdue e-Pubs (Purdue University System) · Cited by 1
GPU acceleration of preconditioned solvers for ill-conditioned linear systems
Gupta, R. (2015) · Research Repository (Delft University of Technology) · doi:10.4233/uuid:ed517292-e790-40b4-a689-8f634d15aa81 · Cited by 1
Adaptive tuning of sparse matrix-vector multiplication on Cell architecture
Cao, Q., Zhao, C., Zhang, Y., Chen, J., Zhu, Y., & He, H. (2010) · International Conference on Computer Engineering and Technology · doi:10.1109/iccet.2010.5485581 · Cited by 1
“The proposal in [22] shows that explicit DMA transfers can be an effective tool, allowing Cell to a very high fraction of the available memory bandwidth.”
BackgroundMethodInfluentialEvaluation of the SUN UltraSparc T2+ Processor for Computational Science
Sandrieser, M., Sabri Pllana, A., & Siegfried Benkner, A. (2009) · Lecture notes in computer science · doi:10.1007/978-3-642-01970-8_97 · Cited by 1
“In [19], a sparse matrix-vector multiply kernel is optimised for different multicore platforms.”
BackgroundSteps Toward Simplifying Sparse Matrix Data Structures
Dinkins, S., Kreaseck, B., & Strout, M. M. (n.d.) · Cited by 1
Sengupta, D. (2016) · SMARTech Repository (Georgia Institute of Technology) · Cited by 1
Davis, E. C. (2020) · ScholarWorks (Boise State University) · doi:10.18122/td/1654/boisestate · Cited by 1
“This is known as the coordinate (COO) format [89].”
BackgroundA parallel fill estimation algorithm for sparse matrices and tensors in blocked formats
Ahrens, W. (2019) · DSpace@MIT (Massachusetts Institute of Technology) · Cited by 1
Worley, P. H. (2010) · Chapman & Hall/CRC computational science series/Chapman & Hall/CRC computational science · doi:10.1201/b10509-18 · Cited by 1
AdaBool: Sparse Boolean Matrix Processing with Adaptive Data RepresentationPreprint
Yang, X., Lei, Y., Yang, J., Zhan, Y., Zheng, Y., Mao, L., Wu, Z., Huang, Y., & Qiu, Y. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.21264 · Cited by 1
“We further evaluate AdaBool on three real-world public datasets— Email-Enron [40], pdb1HYS [60], and Human-Gene1 [5]—spanning nearly three orders of magnitude in density and size.”
Sparse Matrix-Vector Multiplication on GPU
Ashari, A. (2014) · OhioLink ETD Center (Ohio Library and Information Network)
Bramas, B., Coulaud, O., & Sylvand, G. (2015) · Parallel Computing · doi:10.1016/j.parco.2015.07.005
Programmability Analysis of Image Processing Kernels on Multi-/Many-core Platforms
Hemminga, S. (2011) · Research Repository (Delft University of Technology)
Technique detection software for Sparse MatricesPreprint
Khan, M. T., & Usman, A. (2012) · arXiv (Cornell University) · doi:10.48550/arxiv.1202.5964
García-Costa, C. (2014) · Digital Repository (Polytechnic University of Cartagena) · doi:10.31428/10317/4010
Yamassaki, R. T. (2014) · Digital Library of Theses and Dissertations (Universidade de São Paulo) · doi:10.11606/t.3.2014.tde-24122014-120113
Tackling Cache-Line Stealing Effects Using Run-Time Adaptation
Zuckerman, S., & Jalby, W. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-19595-2_5
“Other benchmark results for various platforms on sparse matrix-vector product [26] show the impact of prefetching on such operations.”
BackgroundDynamic Load Balancing of Matrix-Vector Multiplications on Roadrunner Compute Nodes
Sancho, J. C., & Kerbyson, D. J. (2009) · Lecture notes in computer science · doi:10.1007/978-3-642-03869-3_19
Accelerating Similarly Structured Data
Wu, L. K. (2014) · doi:10.7916/d8pr7t46
A performance comparison of linear algebra libraries for sparse matrix-vector product
Augusto, D. A., Carvalho, L. M., Goldfeld, P., Muritiba, A. E. F., & Souza, M. (2015) · Proceeding Series of the Brazilian Society of Computational and Applied Mathematics · doi:10.5540/03.2015.003.01.0116
“1 Benchmark problems All measurements were taken using a set of benchmark square matrices as compiled in [5], covering a variety of fields.”
MethodSibai, F. N., Saad, M., & Kidwai, H. K. (2011) · Concurrency and Computation Practice and Experience · doi:10.1002/cpe.1809
“The computation-demanding elements of CG, precisely the inner products and matrix vector operations, can be directly parallelized [9], and as a result, researchers (e.g. [10]) have mainly focused on analyzing the parallelization of sparse matrix–vector multiplication (SpMV) [11].”
BackgroundDATA ACCESS CHARACTERISTICS AND OPTIMIZATIONS FOR SUN ULTRASPARC T2 AND T2+ SYSTEMS
Hager, G., Zeiser, T., & Wellein, G. (2008) · Parallel Processing Letters · doi:10.1142/s0129626408003521
“It has been shown, however, that kernels which are almost exclusively dominated by loads can achieve somewhat larger bandwidths [4], which leads to the conclusion that at least part of the problem is caused by overhead for bidirectional transfers.”
BackgroundLocality-Aware Dynamic Mapping for Multithreaded Applications
Demiröz, B., Topcuoglu, H. R., Kandemir, M., & Tosun, O. (2012) · Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP) · doi:10.1109/pdp.2012.84
A Modified Storage Format for Accelerating SPMV
Tian, J. W., Wu, F. Q., Zeng, G., & Gong, L. (2014) · International Joint Conference on Computational Sciences and Optimization (CSO) · doi:10.1109/cso.2014.106
Abed, K. H., & Morris, G. R. (2012) · Journal of Parallel and Distributed Computing · doi:10.1016/j.jpdc.2012.07.011
“In a more contemporary research effort, Williams et al. consider techniques to improve the performance of sparse matrix vector multiply on the new multicore GPPs [ 9 ] .”
BackgroundLessons Learnt Porting Parallelisation Techniques for Irregular Codes to NUMA Systems
Lorenzo, J. A., Pichel, J. C., LaFrance-Linden, D., Rivera, F. F., & Singh, D. E. (2010) · Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP) · doi:10.1109/pdp.2010.66
“Likewise, Williams et al. [5] propose several optimisation techniques for the sparse matrix-vector multiplication which are evaluated on different multi-core platforms.”
BackgroundAn Optimized GP-GPU Warp Scheduling Algorithm for Sparse Matrix-Vector Multiplication
Liu, L., Liu, M., & Wang, C. (2013) · IEEE International Conference on Networking, Architecture and Storage (NAS) · doi:10.1109/nas.2013.35
“Sparse matrices can be stored in Diagonal Formats (DIA), Compressed Sparse Row Format (CSR), Compressed Sparse Column Format (CSC), and Block Compressed Sparse Row Format (BCSR) in order to explore the best possible use of the GP-GPU for a variety of algorithmic parameters [3], [5], [10], [14], [17], [18].”
BackgroundMethodAsymmetric flow control for data transfer in hybrid computing systems
Iorio, F., Muller, K., Castelfranco, A., Callanan, O., & Sanuki, T. (2010) · IBM Journal of Research and Development · doi:10.1147/jrd.2009.2036385
Generalized Matrix Multiplication and its Object Oriented Model
Ganzha, M., Paprzycki, M., & Sedukhin, S. G. (2014) · Scalable Computing Practice and Experience · doi:10.12694/scpe.v15i2.983
Runtime program generation and empirical optimization for sparse matrix-vector multiplication
Aktemur, B., Yildiz, A., & Kamin, S. (2012) · Signal Processing and Communications Applications Conference (SIU) · doi:10.1109/siu.2012.6204772
Interaction List Compression in Large Parallel Particle Simulations on Multicore Systems
Rünger, G., & Schwind, M. A. (2012) · Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP) · doi:10.1109/pdp.2012.74
“In [10], the sparse matrix is stored by the common compressed sparse row format using memory efficient indices, which use fewer bytes to address matrix data.”
MethodThe Storage Formats for Accelerating SMVP on a GPU
Tian, J. W., Wu, F. Q., Zou, R., Zeng, G., & Gong, L. (2013) · International Conference on Business Intelligence and Financial Engineering · doi:10.1109/bife.2013.107
“2 CSR The most often used sparse storage format is compressed sparse row (CSR) format([6]).”
BackgroundUPCBLAS : a numerical library for unified parallel C with architecture-aware optimizations
González‐Domínguez, J. (2013) · LA Referencia (Red Federada de Repositorios Institucionales de Publicaciones Científicas)
LDRD final report : managing shared memory data distribution in hybrid HPC applications.
Merritt, A. M., & Pedretti, K. T. (2010) · OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · doi:10.2172/1007320
ANALYSIS OF ASSOCIATIVITY AND CONFLICT MISSES
Gušev, M., & Ristov, S. (2017) · Contributions Section of Natural Mathematical and Biotechnical Sciences · doi:10.20903/csnmbs.masa.2013.34.1-2.42
“The idea for padding the data to reshuffle the data pattern will amortize the performance drawback, since it will avoid associativity problem [5].”
BackgroundTUNED AND ASYNCHRONOUS STENCIL KERNELS FOR CPU/GPU SYSTEMS
Venkatasubramanian, S. (2009) · SMARTech Repository (Georgia Institute of Technology)
Parallel Optimisation Strategies for Fusion Codes
Jackson, A., Reid, F., Booth, S., Hein, J., Westerholm, J., Aspnäs, M., Catala, M., & Soba, A. (2011) · Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP) · doi:10.1109/pdp.2011.15
Chuang, C.-C., Chiu, Y.-S., Kao, Q.-L., Chen, Z.-H., & Lee, C. (2012) · IEEE International Conference on Cloud Computing Technology and Science · doi:10.1109/cloudcom.2012.6427564
“In [12], Lee proposed the block checkerboard method to improve the performance and accuracy of the checkerboard method.”
MethodTowards a high performance parallel library to compute fluid flexible structures interactions
Nagar, P. (2015) · Purdue e-Pubs (Purdue University System) · doi:10.7912/c2/2324
Optimizing Sparse Matrix-Matrix Multiplication on a Heterogeneous CPU-GPU Platform
Wu, X. (2015) · Digital Archive @ GSU · doi:10.57709/7906679
High Performance Adaptive Finite Element Methods
Jansson, N. (2013)
Compressing Method of NetCDF Files Containing Clustered Sparse Matrix
Hwang, S., Choi, G., & Heo, D. (2015) · Lecture notes in electrical engineering · doi:10.1007/978-3-662-47895-0_12
Subdivision Surface Evaluation as Sparse Matrix-Vector Multiplication
Driscoll, M. (2015)
Integrated Nanophotonic Silicon Devices for Next Generation Computing Chips
Djordjevic, S. S. (2014)
On the Complexity of Robust Source-to-Source Translation from CUDA to OpenCL
Sathre, P. (2013) · VTechWorks (Virginia Tech)
A Hardware Multithreaded SpMV Kernel for the Convey HC-2ex
Halstead, R. J., & Najjar, W. A. (2013)
On the design of architecture-aware algorithms for emerging applications
Bader, D. A., & Kang, S. (2011) · SMARTech Repository (Georgia Institute of Technology)
MEMORY EFFICIENCY IMPLICATIONS ON SPARSE MATRIX OPERATIONS
Jain, S., & NC DOCKS at The University of North Carolina at Charlotte (2014) · NC Digital Online Collection of Knowledge and Scholarship (The University of North Carolina at Greensboro)
State of the Art Distributed Parallel Computational Techniques in Industrial Finite Element Analysis
Komzsik, L. (2011) · Computational science, engineering and technology series · doi:10.4203/csets.27.1
FPGA Coprocessor for Simulation of Neural Networks Using Compressed Matrix Storage
Bornschein, J. (2011) · IGI Global eBooks · doi:10.4018/978-1-60960-018-1.ch011
DSLs and Search for Linear Algebra Performance Optimization
Nelson, T. (2015) · CU Scholar (University of Colorado Boulder)
Statistical analysis of industrial processes using fast nonparametric regression techniques
Diehl, S. (2014) · edoc Publication server (Humboldt University of Berlin) · doi:10.18452/14212
“Hence, we introduce only a small additional error if we set these values equal zero but benefit from the sparsity of Φk since tailored methods for sparse matrix multiplication are applicable (cf. Yuster and Zwick (2005), Bell and Garland (2008) and Williams et al. (2009) for furhter readings).”
BackgroundParallel matrix algorithm autotuner on multi-core architecture
Ji-Lin, Z., Yong-Jian, R., Zhang, W., Xu, X., Wan, J., & Jiang-Hui, D. (2010) · International Conference on Information Science and Engineering · doi:10.1109/icise.2010.5691457
Structure-based Optimizations for Sparse Matrix-Vector Multiply
Belgin, M. (2010) · VTechWorks (Virginia Tech)
An integrated compiler and runtime framework for sparse matrix codes
Venkat, A. (2018) · J. Willard Marriott Library · doi:10.26053/0h-n2q9-he00
EVALUATION OF SPARSE LINEAR ALGEBRA OPERATIONS IN TRILINOS
Siahatgar, M., & Voigt, G. V. (2016) · VII European Congress on Computational Methods in Applied Sciences and Engineering · doi:10.7712/100016.1893.11500
“Numerous efforts [1-6] have been made to implement these kernels with optimized algorithms and tuned to specific memory and processor architectures.”
MethodFujimoto, R., Hunter, M. P., & Park, H. (2015) · Defense Technical Information Center (DTIC) · doi:10.21236/ad1004753
“Other research examines issues such as predictive modeling of energy and power (Czechowski and Vuduc 2013;Williams et al. 2009; Keckler et al. 2011) and embedded systems evaluations (Grasso, et al.”
BackgroundData parallel algebraic multigrid
Dalton, S. (2015) · Illinois Digital Environment for Access to Learning and Scholarship (University of Illinois at Urbana-Champaign) · Cited by 1
Infraestructura para el análisis de rendimiento
More, A. (2017) · El Servicio de Difusión de la Creación Intelectual (National University of La Plata) · doi:10.35537/10915/59339
White-box methodologies, programming abstractions and librariesPreprint
Ha, P. H., Tran, N. N. V., Umar, I., Atalar, A., Gidenstam, A., Renaud-Goud, P., & Tsigas, P. (2015) · arXiv (Cornell University)
Ha, P. H., Tran, V. N.-N., Umar, I., Atalar, A., Gidenstam, A., Renaud-Goud, P., Tsigas, P., & Walulya, I. (2016) · arXiv (Cornell University)
Etude de la Distribution de Calculs Creux sur une Grappe Multi-coeursPreprint
Ayachi, M. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1902.02156
Parallelism, patterns, and performance in iterative mri reconstruction
Lustig, M. S., Keutzer, K., & Murphy, M. (2011)
Modeling Energy Consumption of Computing Systems: from Homogeneous to Heterogeneous Systems
Tran, V. N.-N. (2019) · NORA - Norwegian Open Research Archives
Duff Aygıtı Tabanlı Seyrek Matris-Vektör Çarpımı
Aktemur, B. (2020) · Deu Muhendislik Fakultesi Fen ve Muhendislik · doi:10.21205/deufmd.2020226501
“[1] Williams S, Oliker L, Vuduc R, Shalf J, Yelick K, Demmel J.”
BackgroundJoint direct and transposed sparse matrix‐vector multiplication for multithreaded CPUs
Kozický, C., & Šimeček, I. (2021) · Concurrency and Computation Practice and Experience · doi:10.1002/cpe.6236
Architecting Memory Systems for Emerging Technologies
Oh, B. (2018) · Deep Blue (University of Michigan)
Optimized Sparse Matrix Operations and Hardware Implementation using FPGA
Murthy, D. K. (2021)
A New Approach to Deep-Learning Model Sparsity via Tensor-with-Sparsity-Attribute
Zheng, N., Lin, B., Zhang, Q., Ma, L., Yang, Y., Yang, F., Yang, M., & Zhou, L. (2021) · Cited by 6
Sparse Matrix-Vector Multiplication Cache Performance Evaluation and Design Exploration
Cui, J., Lü, K., & Liu, S. (2021) · Proceedings - International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems · doi:10.1109/mascots53633.2021.9614301
Composing Loop-carried Dependence with Other LoopsPreprint
Cheshmi, K., Strout, M. M., & Dehnavi, M. M. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2111.12238
“For example, the sparse triangular solve has been optimized in [28, 35, 37, 38, 43, 53, 55, 57, 59], optimizations of sparse matrix-vector multiply are available in [3, 23, 27, 30, 32, 58], and LU and Cholesky factorization have been optimized in SuperLU [29] and Pastix [19].”
Differentiating-based Vectorization for Sparse KernelsPreprint
Cetinic, Z., Cheshmi, K., & Dehnavi, M. M. (2021) · arXiv (Cornell University)
Performance Optimization for Parallel SpMV on a NUMA Architecture
Zhang, Y., Yang, W., Li, K., & Cai, Q. (2022) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-3-030-89698-0_131
Schüle, J. (2010) · Oldenbourg Wissenschaftsverlag eBooks · doi:10.1524/9783486711585.bm
HICAMP: architectural support for efficient concurrency-safe shared structured data access
Cheriton, D. R., Firoozshahian, A., Solomatnikov, A., Stevenson, J. P., & Azizi, O. (2012) · ACM SIGARCH Computer Architecture News · doi:10.1145/2189750.2151007
AUXÍLIO A PORTABILIDADE DE CÓDIGO EM APLICAÇÕES DE ALTO DESEMPENHO
FILHO, P. R. P. D. S. (2016) · doi:10.17771/pucrio.acad.28707
GPUBLQMR: GPU-accelerated sparse block quasi-minimum residual linear solver
Lacouture, R. (2021) · Digital Repository Service (Northeastern University) · doi:10.17760/d20410372
Exploring high performance sparse operations on GPUs
Karimi, E. (2022) · Digital Repository Service (Northeastern University) · doi:10.17760/d20471051
SpMV Profiling and Optimization Analysis
Journals, I., Kulkarni, A. V., & Barde, C. R. (2015) · Figshare · doi:10.6084/m9.figshare.1495296
A Hybrid Machine Learning Method for Cross-Platform Performance Prediction of Parallel Applications
Mahdavi, K. (2024) · Proceedings of the International Conference on Parallel Processing · doi:10.1145/3673038.3673059
“Their model provide useful insights into locality effects but required detailed system knowledge [24].”
SpDCache: Region-Based Reduction Cache for Outer-Product Sparse Matrix Kernels
Isaac--Chassande, V., Evans, A., Durand, Y., & Rousseau, F. (2024) · IEEE International Conference on Application-specific Systems, Architectures and Processors (ASAP) · doi:10.1109/asap61560.2024.00012
Exploration of Calculation Methods for a Class of Sparse Determinants
Xia, D., Yan, Y., & Zhou, J. (2024) · JJournal of natural science education. · doi:10.62517/jnse.202417508
Simonsen, L.-I. H. (2013) · BIBSYS Brage (BIBSYS (Norway))
Engineering of Algorithms for Very Large k Partitioning
Haag, M. (2021) · Repository KITopen (Karlsruhe Institute of Technology) · doi:10.5445/ir/1000150361
CUDA OPTIMIZATION OF A CLASS OF ITERATIVE LINEAR ALGEBRA SOLVERS
Firdaus, L. (2017) · Research Publication Repository of King Fahd University of Petroleum and Minerals (King Fahd University of Petroleum and Minerals)
Benchmarking mixed-mode PETSc performance on high-performance architecturesPreprint
Lange, M., Gorman, G., Weiland, M., Mitchell, L., Guo, X., & Southern, J. A. (2013) · arXiv (Cornell University) · doi:10.48550/arxiv.1307.4567
“Instead, thread-level load balance may be improved statically by dividing the number of non-zeros approximately equally between threads, as pointed out by Williams et al. [7].”
BackgroundTaco Tensor Algebra kernels on distributed systems using Legion
Shinde, S. D. (2019) · DSpace@MIT (Massachusetts Institute of Technology)
OVERT: Orchestrating Vector-Scalar Execution for Efficient SpMV on Modern CPUs
Lei, K., Yang, H., Zhang, K., Du, S., Casas, M., Xu, Y., Luan, Z., Liu, Y., & Qian, D. (2025) · Proceedings of the International Conference on Parallel Processing · doi:10.1145/3754598.3754632
A multi-level programming paradigm for numerical computing on post-petascale and exascale machines
Hugues, M. (2011) · theses.fr (ABES) · doi:10.70675/96f23fa3z3396z4802z8a52z0c0a7ae008f5
“These sparse matrices were also used in the paper [43] to compare optimization and performance of SpMV on different multicore processors.”
Contributions à l'algèbre linéaire exacte sur corps finis et au chiffrement homomorphe
Vialla, B. (2015) · theses.fr (ABES) · doi:10.70675/d3207967z5bcbz4795z9f83z0219aae207d5
Architectures matérielles numériques intégrées et réseaux de neurones à codage parcimonieux
Wouafo, H. N. (2016) · theses.fr (ABES) · doi:10.70675/a4cc5e39zf913z41ebz9b82zbb678b263f5e
“Plusieurs implémentations ont été proposées pour mettre en œuvre des SpMxV que ce soit sur CPU (Central Processing Unit) [88], FPGA [89]–[92] ou encore sur GPU (Graphics Processing Unit) [93]–[95].”
BackgroundAccélérateurs logiciels et matériels pour l'algèbre linéaire creuse sur les corps finis
Jeljeli, H. (2015) · theses.fr (ABES) · doi:10.70675/cc0f4d46z15bez41e2za6c5z3c3f5f801ab9
Auto-tuning pour la détection automatique du meilleur format de compression pour matrice creuse
Mehrez, I. (2018) · theses.fr (ABES) · doi:10.70675/80781fb1z7a4fz4b2bz99a9zcbf5282669e0
Metadata
- Authors
- Williams, S., Oliker, L., Vuduc, R., Shalf, J., Yelick, K., & Demmel, J.
- Affiliations
- Lawrence Berkeley National Laboratory (US)
- Georgia Institute of Technology (US)
- University of California, Berkeley (US)
- Corresponding author
- Williams, Samuel
- Venue
- Proceedings of the 2007 ACM/IEEE conference on Supercomputing
- Publisher
- ACM
- Published
- 2007-11-10
- Volume
- 35
- Issue
- 3
- Pages
- 1-12
- ISSN
- 0167-8191 (ISSN-L)
- OpenAlex
- W1527464046
- Cited by
- 791
- References
- 23
- Citations by year
- 2026: 7 · 2025: 13 · 2024: 14 · 2023: 17 · 2022: 20 · 2021: 34 · 2020: 27 · 2019: 41 · 2018: 41 · 2017: 33 · 2016: 57 · 2015: 56 · 2014: 78 · 2013: 66 · 2012: 63
- FWCI
- 93.97
- Type
- journal article
- Language
- en
- Open access
- green
- escholarship.org
- Locations
- Parallel Computing (publishedVersion)
- ACM/IEEE Conference on Supercomputing (SC) (publishedVersion)
- eScholarship (California Digital Library) (submittedVersion, cc-by)
- eScholarship (California Digital Library) (submittedVersion)
- OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion)
- University of North Texas Digital Library (University of North Texas) (submittedVersion)
- Lawrence Berkeley National Laboratory
- OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information)
- Funders
- Office of Science (DE-AC02-05CH11231)
- Advanced Scientific Computing Research (DE-AC02-05CH11231)
- Topics
- Physical Sciences › Computer Science › Hardware and Architecture › Parallel Computing and Optimization Techniques
- Physical Sciences › Computer Science › Computational Theory and Mathematics › Matrix Theory and Algorithms
- Physical Sciences › Computer Science › Computer Networks and Communications › Distributed and Parallel Computing Systems
- Keywords
- sparse matrix-vector multiplication, multi-core processors, Cell processor, multithreading
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.