Matches in SemOpenAlex for { <https://semopenalex.org/work/W2607112060> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W2607112060 endingPage "56" @default.
- W2607112060 startingPage "37" @default.
- W2607112060 abstract "Abstract Typically, the execution time of a kernel on a GPU is a difficult to predict measure as it depends on a wide range of factors. Performance can be limited by either memory transfer, compute throughput or other latencies. In this paper, we improve on the roofline model following a quantitative approach and present a completely automated GPU performance prediction technique. In this respect this model utilizes micro-benchmarking and profiling in a “black box” fashion as no inspection of source/binary code is required. The proposed model combines parameters in order to characterize the performance limiting factor and to estimate execution time. In addition, we propose the quadrant-split visual representation, which captures the characteristics of multiple processors in relation to a particular kernel. We performed experiments on stencil computation (red/black SOR), SGEMM and a total of 28 kernels of the Rodinia benchmark suite, using six CUDA GPUs and we showed an absolute error in predictions of 27.66% in the average case. Furthermore, the performance model was also examined on an AMD GPU through the HIP programming environment. Prediction errors were comparable despite the significant architectural differences between different vendor GPUs." @default.
- W2607112060 created "2017-04-28" @default.
- W2607112060 creator A5047221264 @default.
- W2607112060 creator A5071807859 @default.
- W2607112060 date "2017-09-01" @default.
- W2607112060 modified "2023-10-18" @default.
- W2607112060 title "A quantitative roofline model for GPU kernel performance estimation using micro-benchmarks and hardware metric profiling" @default.
- W2607112060 cites W1588403183 @default.
- W2607112060 cites W1904624772 @default.
- W2607112060 cites W1976139981 @default.
- W2607112060 cites W1981557297 @default.
- W2607112060 cites W1984296775 @default.
- W2607112060 cites W2002555321 @default.
- W2607112060 cites W2035276889 @default.
- W2607112060 cites W2038666141 @default.
- W2607112060 cites W2060076314 @default.
- W2607112060 cites W2063750261 @default.
- W2607112060 cites W2074084090 @default.
- W2607112060 cites W2080592089 @default.
- W2607112060 cites W2082238959 @default.
- W2607112060 cites W2105545110 @default.
- W2607112060 cites W2106821446 @default.
- W2607112060 cites W2128912557 @default.
- W2607112060 cites W2130336316 @default.
- W2607112060 cites W2140348470 @default.
- W2607112060 cites W2153495590 @default.
- W2607112060 cites W2153667821 @default.
- W2607112060 cites W2276930717 @default.
- W2607112060 cites W2306144962 @default.
- W2607112060 cites W2311803642 @default.
- W2607112060 cites W2417350949 @default.
- W2607112060 cites W2997701623 @default.
- W2607112060 cites W4250047567 @default.
- W2607112060 doi "https://doi.org/10.1016/j.jpdc.2017.04.002" @default.
- W2607112060 hasPublicationYear "2017" @default.
- W2607112060 type Work @default.
- W2607112060 sameAs 2607112060 @default.
- W2607112060 citedByCount "38" @default.
- W2607112060 countsByYear W26071120602018 @default.
- W2607112060 countsByYear W26071120602019 @default.
- W2607112060 countsByYear W26071120602020 @default.
- W2607112060 countsByYear W26071120602021 @default.
- W2607112060 countsByYear W26071120602022 @default.
- W2607112060 countsByYear W26071120602023 @default.
- W2607112060 crossrefType "journal-article" @default.
- W2607112060 hasAuthorship W2607112060A5047221264 @default.
- W2607112060 hasAuthorship W2607112060A5071807859 @default.
- W2607112060 hasConcept C118615104 @default.
- W2607112060 hasConcept C121684516 @default.
- W2607112060 hasConcept C162324750 @default.
- W2607112060 hasConcept C173608175 @default.
- W2607112060 hasConcept C176217482 @default.
- W2607112060 hasConcept C187191949 @default.
- W2607112060 hasConcept C199360897 @default.
- W2607112060 hasConcept C21547014 @default.
- W2607112060 hasConcept C33923547 @default.
- W2607112060 hasConcept C41008148 @default.
- W2607112060 hasConcept C74193536 @default.
- W2607112060 hasConceptScore W2607112060C118615104 @default.
- W2607112060 hasConceptScore W2607112060C121684516 @default.
- W2607112060 hasConceptScore W2607112060C162324750 @default.
- W2607112060 hasConceptScore W2607112060C173608175 @default.
- W2607112060 hasConceptScore W2607112060C176217482 @default.
- W2607112060 hasConceptScore W2607112060C187191949 @default.
- W2607112060 hasConceptScore W2607112060C199360897 @default.
- W2607112060 hasConceptScore W2607112060C21547014 @default.
- W2607112060 hasConceptScore W2607112060C33923547 @default.
- W2607112060 hasConceptScore W2607112060C41008148 @default.
- W2607112060 hasConceptScore W2607112060C74193536 @default.
- W2607112060 hasLocation W26071120601 @default.
- W2607112060 hasOpenAccess W2607112060 @default.
- W2607112060 hasPrimaryLocation W26071120601 @default.
- W2607112060 hasRelatedWork W1507231479 @default.
- W2607112060 hasRelatedWork W1580730938 @default.
- W2607112060 hasRelatedWork W1604898313 @default.
- W2607112060 hasRelatedWork W2348361596 @default.
- W2607112060 hasRelatedWork W2624440775 @default.
- W2607112060 hasRelatedWork W2792115777 @default.
- W2607112060 hasRelatedWork W2998249817 @default.
- W2607112060 hasRelatedWork W4230635966 @default.
- W2607112060 hasRelatedWork W4250047567 @default.
- W2607112060 hasRelatedWork W2521117258 @default.
- W2607112060 hasVolume "107" @default.
- W2607112060 isParatext "false" @default.
- W2607112060 isRetracted "false" @default.
- W2607112060 magId "2607112060" @default.
- W2607112060 workType "article" @default.