Matches in SemOpenAlex for { <https://semopenalex.org/work/W1839773802> ?p ?o ?g. }
Showing items 1 to 93 of
93
with 100 items per page.
- W1839773802 endingPage "31" @default.
- W1839773802 startingPage "1" @default.
- W1839773802 abstract "KBLAS is an open-source, high-performance library that provides optimized kernels for a subset of Level 2 BLAS functionalities on CUDA-enabled GPUs. Since performance of dense matrix-vector multiplication is hindered by the overhead of memory accesses, a double-buffering optimization technique is employed to overlap data motion with computation. After identifying a proper set of tuning parameters, KBLAS efficiently runs on various GPU architectures while avoiding code rewriting and retaining compliance with the standard BLAS API. Another optimization technique allows ensuring coalesced memory access when dealing with submatrices, especially for high-level dense linear algebra algorithms. All KBLAS kernels have been leveraged to a multi-GPU environment, which requires the introduction of new APIs. Considering general matrices, KBLAS is very competitive with existing state-of-the-art kernels and provides a smoother performance across a wide range of matrix dimensions. Considering symmetric and Hermitian matrices, the KBLAS performance outperforms existing state-of-the-art implementations on all matrix sizes and achieves asymptotically up to 50% and 60% speedup against the best competitor on single GPU and multi-GPUs systems, respectively. Performance results also validate our performance model. A subset of KBLAS high-performance kernels have been integrated into NVIDIA's standard BLAS implementation (cuBLAS) for larger dissemination, starting from version 6.0." @default.
- W1839773802 created "2016-06-24" @default.
- W1839773802 creator A5017526753 @default.
- W1839773802 creator A5021283893 @default.
- W1839773802 creator A5077268543 @default.
- W1839773802 date "2016-05-10" @default.
- W1839773802 modified "2023-09-24" @default.
- W1839773802 title "KBLAS" @default.
- W1839773802 cites W1473251595 @default.
- W1839773802 cites W1957031661 @default.
- W1839773802 cites W2002555321 @default.
- W1839773802 cites W2016279572 @default.
- W1839773802 cites W2048558763 @default.
- W1839773802 cites W2063186542 @default.
- W1839773802 cites W2090593986 @default.
- W1839773802 cites W2099021415 @default.
- W1839773802 cites W2125960020 @default.
- W1839773802 cites W4241513866 @default.
- W1839773802 doi "https://doi.org/10.1145/2818311" @default.
- W1839773802 hasPublicationYear "2016" @default.
- W1839773802 type Work @default.
- W1839773802 sameAs 1839773802 @default.
- W1839773802 citedByCount "29" @default.
- W1839773802 countsByYear W18397738022015 @default.
- W1839773802 countsByYear W18397738022016 @default.
- W1839773802 countsByYear W18397738022017 @default.
- W1839773802 countsByYear W18397738022018 @default.
- W1839773802 countsByYear W18397738022019 @default.
- W1839773802 countsByYear W18397738022020 @default.
- W1839773802 countsByYear W18397738022021 @default.
- W1839773802 countsByYear W18397738022022 @default.
- W1839773802 countsByYear W18397738022023 @default.
- W1839773802 crossrefType "journal-article" @default.
- W1839773802 hasAuthorship W1839773802A5017526753 @default.
- W1839773802 hasAuthorship W1839773802A5021283893 @default.
- W1839773802 hasAuthorship W1839773802A5077268543 @default.
- W1839773802 hasConcept C111919701 @default.
- W1839773802 hasConcept C114614502 @default.
- W1839773802 hasConcept C121332964 @default.
- W1839773802 hasConcept C139352143 @default.
- W1839773802 hasConcept C17349429 @default.
- W1839773802 hasConcept C173608175 @default.
- W1839773802 hasConcept C21442007 @default.
- W1839773802 hasConcept C2524010 @default.
- W1839773802 hasConcept C2778119891 @default.
- W1839773802 hasConcept C2779960059 @default.
- W1839773802 hasConcept C2780595030 @default.
- W1839773802 hasConcept C33923547 @default.
- W1839773802 hasConcept C41008148 @default.
- W1839773802 hasConcept C50630238 @default.
- W1839773802 hasConcept C62520636 @default.
- W1839773802 hasConcept C68339613 @default.
- W1839773802 hasConcept C83283714 @default.
- W1839773802 hasConcept C84114770 @default.
- W1839773802 hasConceptScore W1839773802C111919701 @default.
- W1839773802 hasConceptScore W1839773802C114614502 @default.
- W1839773802 hasConceptScore W1839773802C121332964 @default.
- W1839773802 hasConceptScore W1839773802C139352143 @default.
- W1839773802 hasConceptScore W1839773802C17349429 @default.
- W1839773802 hasConceptScore W1839773802C173608175 @default.
- W1839773802 hasConceptScore W1839773802C21442007 @default.
- W1839773802 hasConceptScore W1839773802C2524010 @default.
- W1839773802 hasConceptScore W1839773802C2778119891 @default.
- W1839773802 hasConceptScore W1839773802C2779960059 @default.
- W1839773802 hasConceptScore W1839773802C2780595030 @default.
- W1839773802 hasConceptScore W1839773802C33923547 @default.
- W1839773802 hasConceptScore W1839773802C41008148 @default.
- W1839773802 hasConceptScore W1839773802C50630238 @default.
- W1839773802 hasConceptScore W1839773802C62520636 @default.
- W1839773802 hasConceptScore W1839773802C68339613 @default.
- W1839773802 hasConceptScore W1839773802C83283714 @default.
- W1839773802 hasConceptScore W1839773802C84114770 @default.
- W1839773802 hasIssue "3" @default.
- W1839773802 hasLocation W18397738021 @default.
- W1839773802 hasOpenAccess W1839773802 @default.
- W1839773802 hasPrimaryLocation W18397738021 @default.
- W1839773802 hasRelatedWork W126146189 @default.
- W1839773802 hasRelatedWork W1501847821 @default.
- W1839773802 hasRelatedWork W1508286210 @default.
- W1839773802 hasRelatedWork W2105812312 @default.
- W1839773802 hasRelatedWork W2144511445 @default.
- W1839773802 hasRelatedWork W2392023973 @default.
- W1839773802 hasRelatedWork W2952625562 @default.
- W1839773802 hasRelatedWork W3038415719 @default.
- W1839773802 hasRelatedWork W3176814699 @default.
- W1839773802 hasRelatedWork W52302056 @default.
- W1839773802 hasVolume "42" @default.
- W1839773802 isParatext "false" @default.
- W1839773802 isRetracted "false" @default.
- W1839773802 magId "1839773802" @default.
- W1839773802 workType "article" @default.