Matches in SemOpenAlex for { <https://semopenalex.org/work/W2891818448> ?p ?o ?g. }
- W2891818448 endingPage "27" @default.
- W2891818448 startingPage "1" @default.
- W2891818448 abstract "The Sparse Matrix-Vector Multiplication (SpMV) kernel dominates the computing cost in numerous scientific applications. Many implementations based on different sparse formats were proposed to improve this kernel on the recent GPU architectures. However, it has been widely observed that there is no “best-for-all” sparse format for the SpMV kernel on GPU. Indeed, serious performance degradation of an order of magnitude can be observed without a careful selection of the sparse format to use. To address this problem, we propose in this article BestSF (Best Sparse Format), a new learning-based sparse meta-format that automatically selects the most appropriate sparse format for a given input matrix. To do so, BestSF relies on a cost-sensitive classification system trained using Weighted Support Vector Machines (WSVMs) to predict the best sparse format for each input sparse matrix. Our experimental results on two different NVIDIA GPU architectures using a large number of real-world sparse matrices show that BestSF achieved a noticeable overall performance improvement over using a single sparse format. While BestSF is trained to select the best sparse format in terms of performance (GFLOPS), our further experimental investigations revealed that using BestSF also led, in most of the test cases, to the best energy efficiency (MFLOPS/W). To prove its practical effectiveness, we also evaluate the performance and energy efficiency improvement achieved when using BestSF as a building block in a GPU-based Preconditioned Conjugate Gradient (PCG) iterative solver." @default.
- W2891818448 created "2018-09-27" @default.
- W2891818448 creator A5035545488 @default.
- W2891818448 creator A5046211358 @default.
- W2891818448 creator A5057958949 @default.
- W2891818448 creator A5084655806 @default.
- W2891818448 date "2018-09-04" @default.
- W2891818448 modified "2023-09-27" @default.
- W2891818448 title "BestSF" @default.
- W2891818448 cites W1471542436 @default.
- W2891818448 cites W1506342804 @default.
- W2891818448 cites W1555620806 @default.
- W2891818448 cites W1588915715 @default.
- W2891818448 cites W1965034778 @default.
- W2891818448 cites W1968120338 @default.
- W2891818448 cites W1971367716 @default.
- W2891818448 cites W1981557297 @default.
- W2891818448 cites W1992476323 @default.
- W2891818448 cites W1993704253 @default.
- W2891818448 cites W2009654791 @default.
- W2891818448 cites W2011159963 @default.
- W2891818448 cites W2017337590 @default.
- W2891818448 cites W2020526771 @default.
- W2891818448 cites W2022219663 @default.
- W2891818448 cites W2023930909 @default.
- W2891818448 cites W2035080386 @default.
- W2891818448 cites W2040556424 @default.
- W2891818448 cites W2045920776 @default.
- W2891818448 cites W2050127041 @default.
- W2891818448 cites W2053060341 @default.
- W2891818448 cites W2086392083 @default.
- W2891818448 cites W2087507944 @default.
- W2891818448 cites W2091883426 @default.
- W2891818448 cites W2095230296 @default.
- W2891818448 cites W2099714261 @default.
- W2891818448 cites W2102528350 @default.
- W2891818448 cites W2104120668 @default.
- W2891818448 cites W2105545110 @default.
- W2891818448 cites W2119821739 @default.
- W2891818448 cites W2147148915 @default.
- W2891818448 cites W2153635508 @default.
- W2891818448 cites W2172000360 @default.
- W2891818448 cites W2479666647 @default.
- W2891818448 cites W2559809597 @default.
- W2891818448 cites W2569190819 @default.
- W2891818448 cites W2570841452 @default.
- W2891818448 cites W2617808919 @default.
- W2891818448 cites W2997701623 @default.
- W2891818448 cites W4240739471 @default.
- W2891818448 cites W4243261006 @default.
- W2891818448 cites W4247712932 @default.
- W2891818448 cites W4249999659 @default.
- W2891818448 cites W610778723 @default.
- W2891818448 cites W2530001924 @default.
- W2891818448 doi "https://doi.org/10.1145/3226228" @default.
- W2891818448 hasPublicationYear "2018" @default.
- W2891818448 type Work @default.
- W2891818448 sameAs 2891818448 @default.
- W2891818448 citedByCount "18" @default.
- W2891818448 countsByYear W28918184482019 @default.
- W2891818448 countsByYear W28918184482020 @default.
- W2891818448 countsByYear W28918184482021 @default.
- W2891818448 countsByYear W28918184482022 @default.
- W2891818448 countsByYear W28918184482023 @default.
- W2891818448 crossrefType "journal-article" @default.
- W2891818448 hasAuthorship W2891818448A5035545488 @default.
- W2891818448 hasAuthorship W2891818448A5046211358 @default.
- W2891818448 hasAuthorship W2891818448A5057958949 @default.
- W2891818448 hasAuthorship W2891818448A5084655806 @default.
- W2891818448 hasBestOaLocation W28918184481 @default.
- W2891818448 hasConcept C11413529 @default.
- W2891818448 hasConcept C114614502 @default.
- W2891818448 hasConcept C121332964 @default.
- W2891818448 hasConcept C124066611 @default.
- W2891818448 hasConcept C163716315 @default.
- W2891818448 hasConcept C173608175 @default.
- W2891818448 hasConcept C199360897 @default.
- W2891818448 hasConcept C2524010 @default.
- W2891818448 hasConcept C2777210771 @default.
- W2891818448 hasConcept C2778770139 @default.
- W2891818448 hasConcept C2780595030 @default.
- W2891818448 hasConcept C33923547 @default.
- W2891818448 hasConcept C3826847 @default.
- W2891818448 hasConcept C41008148 @default.
- W2891818448 hasConcept C459310 @default.
- W2891818448 hasConcept C54219892 @default.
- W2891818448 hasConcept C56372850 @default.
- W2891818448 hasConcept C62520636 @default.
- W2891818448 hasConcept C74193536 @default.
- W2891818448 hasConcept C81184566 @default.
- W2891818448 hasConceptScore W2891818448C11413529 @default.
- W2891818448 hasConceptScore W2891818448C114614502 @default.
- W2891818448 hasConceptScore W2891818448C121332964 @default.
- W2891818448 hasConceptScore W2891818448C124066611 @default.
- W2891818448 hasConceptScore W2891818448C163716315 @default.
- W2891818448 hasConceptScore W2891818448C173608175 @default.
- W2891818448 hasConceptScore W2891818448C199360897 @default.
- W2891818448 hasConceptScore W2891818448C2524010 @default.