Matches in SemOpenAlex for { <https://semopenalex.org/work/W2725757310> ?p ?o ?g. }
- W2725757310 abstract "Because sparse matrix-vector multiplication (SpMV) is an important and widely used computational kernel in many real-world applications, it behooves us to accelerate SpMV on modern multi- and many-core architectures. While many storage formats have been developed to facilitate SpMV operations, the compressed sparse row (CSR) format is still the most popular and general storage format. However, parallelizing CSR-based SpMV on multi- and many-core processors (e.g., CPUs, APUs, GPUs) remains a challenging problem, including dealing with uncoalesced memory access, balancing workload, and identifying the most appropriate parallelizing strategy. In the paper, we propose a novel auto-tuning framework that automatically finds the most efficient parallelizing strategy to achieve high-performance SpMV. Our framework can deter- mine the right binning schemes to group similar workloads into bins (e.g., buckets) with negligible overhead. Then, for each bin, the most suitable kernel is selected to process the rows within. Our framework is input-aware and based on a machine-learning method. The results show that our auto-tuned SpMV performs significantly better than the default SpMV. The speedups on 16 representative matrices range from 1.2x to 52.0x. Compared to the state-of-the-art SpMV kernel, our work yields better performance in most cases, achieving up to a 1.9x speedup." @default.
- W2725757310 created "2017-07-14" @default.
- W2725757310 creator A5018356320 @default.
- W2725757310 creator A5047871879 @default.
- W2725757310 creator A5058539554 @default.
- W2725757310 date "2017-05-01" @default.
- W2725757310 modified "2023-10-06" @default.
- W2725757310 title "Auto-Tuning Strategies for Parallelizing Sparse Matrix-Vector (SpMV) Multiplication on Multi- and Many-Core Processors" @default.
- W2725757310 cites W1980938256 @default.
- W2725757310 cites W1987840949 @default.
- W2725757310 cites W1990832096 @default.
- W2725757310 cites W1993704253 @default.
- W2725757310 cites W2003117296 @default.
- W2725757310 cites W2009654791 @default.
- W2725757310 cites W2014214820 @default.
- W2725757310 cites W2023930909 @default.
- W2725757310 cites W2035080386 @default.
- W2725757310 cites W2053087996 @default.
- W2725757310 cites W2066636486 @default.
- W2725757310 cites W2084446096 @default.
- W2725757310 cites W2087507944 @default.
- W2725757310 cites W2088866486 @default.
- W2725757310 cites W2091883426 @default.
- W2725757310 cites W2095249664 @default.
- W2725757310 cites W2104262506 @default.
- W2725757310 cites W2111377561 @default.
- W2725757310 cites W2115052535 @default.
- W2725757310 cites W2128853364 @default.
- W2725757310 cites W2162283062 @default.
- W2725757310 cites W2186072464 @default.
- W2725757310 cites W2260499435 @default.
- W2725757310 cites W2263374743 @default.
- W2725757310 cites W2347043188 @default.
- W2725757310 cites W2411480360 @default.
- W2725757310 cites W2474844316 @default.
- W2725757310 cites W2480699195 @default.
- W2725757310 cites W2506072259 @default.
- W2725757310 cites W2538301624 @default.
- W2725757310 cites W2559832905 @default.
- W2725757310 cites W4236619648 @default.
- W2725757310 cites W4247828381 @default.
- W2725757310 doi "https://doi.org/10.1109/ipdpsw.2017.155" @default.
- W2725757310 hasPublicationYear "2017" @default.
- W2725757310 type Work @default.
- W2725757310 sameAs 2725757310 @default.
- W2725757310 citedByCount "23" @default.
- W2725757310 countsByYear W27257573102017 @default.
- W2725757310 countsByYear W27257573102018 @default.
- W2725757310 countsByYear W27257573102019 @default.
- W2725757310 countsByYear W27257573102020 @default.
- W2725757310 countsByYear W27257573102021 @default.
- W2725757310 countsByYear W27257573102022 @default.
- W2725757310 countsByYear W27257573102023 @default.
- W2725757310 crossrefType "proceedings-article" @default.
- W2725757310 hasAuthorship W2725757310A5018356320 @default.
- W2725757310 hasAuthorship W2725757310A5047871879 @default.
- W2725757310 hasAuthorship W2725757310A5058539554 @default.
- W2725757310 hasConcept C111919701 @default.
- W2725757310 hasConcept C114614502 @default.
- W2725757310 hasConcept C121332964 @default.
- W2725757310 hasConcept C162324750 @default.
- W2725757310 hasConcept C163716315 @default.
- W2725757310 hasConcept C17349429 @default.
- W2725757310 hasConcept C173608175 @default.
- W2725757310 hasConcept C21547014 @default.
- W2725757310 hasConcept C24890656 @default.
- W2725757310 hasConcept C2778915421 @default.
- W2725757310 hasConcept C2779960059 @default.
- W2725757310 hasConcept C2780595030 @default.
- W2725757310 hasConcept C33923547 @default.
- W2725757310 hasConcept C41008148 @default.
- W2725757310 hasConcept C56372850 @default.
- W2725757310 hasConcept C62520636 @default.
- W2725757310 hasConcept C68339613 @default.
- W2725757310 hasConcept C74193536 @default.
- W2725757310 hasConcept C78766204 @default.
- W2725757310 hasConcept C84114770 @default.
- W2725757310 hasConceptScore W2725757310C111919701 @default.
- W2725757310 hasConceptScore W2725757310C114614502 @default.
- W2725757310 hasConceptScore W2725757310C121332964 @default.
- W2725757310 hasConceptScore W2725757310C162324750 @default.
- W2725757310 hasConceptScore W2725757310C163716315 @default.
- W2725757310 hasConceptScore W2725757310C17349429 @default.
- W2725757310 hasConceptScore W2725757310C173608175 @default.
- W2725757310 hasConceptScore W2725757310C21547014 @default.
- W2725757310 hasConceptScore W2725757310C24890656 @default.
- W2725757310 hasConceptScore W2725757310C2778915421 @default.
- W2725757310 hasConceptScore W2725757310C2779960059 @default.
- W2725757310 hasConceptScore W2725757310C2780595030 @default.
- W2725757310 hasConceptScore W2725757310C33923547 @default.
- W2725757310 hasConceptScore W2725757310C41008148 @default.
- W2725757310 hasConceptScore W2725757310C56372850 @default.
- W2725757310 hasConceptScore W2725757310C62520636 @default.
- W2725757310 hasConceptScore W2725757310C68339613 @default.
- W2725757310 hasConceptScore W2725757310C74193536 @default.
- W2725757310 hasConceptScore W2725757310C78766204 @default.
- W2725757310 hasConceptScore W2725757310C84114770 @default.
- W2725757310 hasLocation W27257573101 @default.
- W2725757310 hasOpenAccess W2725757310 @default.
- W2725757310 hasPrimaryLocation W27257573101 @default.