Matches in SemOpenAlex for { <https://semopenalex.org/work/W2122338824> ?p ?o ?g. }
- W2122338824 endingPage "577" @default.
- W2122338824 startingPage "557" @default.
- W2122338824 abstract "Summary Achieving optimal performance on the latest multi‐core and many‐core architectures increasingly depends on making efficient use of the hardware's vector units. This paper presents results on achieving high performance through vectorization on CPUs and the Xeon‐Phi on a key class of irregular applications: unstructured mesh computations. Using single instruction multiple thread (SIMT) and single instruction multiple data (SIMD) programming models, we show how unstructured mesh computations map to OpenCL or vector intrinsics through the use of code generation techniques in the OP2 Domain Specific Library and explore how irregular memory accesses and race conditions can be organized on different hardware. We benchmark Intel Xeon CPUs and the Xeon‐Phi, using a tsunami simulation and a representative CFD benchmark. Results are compared with previous work on CPUs and NVIDIA GPUs to provide a comparison of achievable performance on current many‐core systems. We show that auto‐vectorization and the OpenCL SIMT model do not map efficiently to CPU vector units because of vectorization issues and threading overheads. In contrast, using SIMD vector intrinsics imposes some restrictions and requires more involved programming techniques but results in efficient code and near‐optimal performance, two times faster than non‐vectorized code. We observe that the Xeon‐Phi does not provide good performance for these applications but is still comparable with a pair of mid‐range Xeon chips. Copyright © 2015 John Wiley & Sons, Ltd." @default.
- W2122338824 created "2016-06-24" @default.
- W2122338824 creator A5035189605 @default.
- W2122338824 creator A5048478302 @default.
- W2122338824 creator A5056186758 @default.
- W2122338824 creator A5070789282 @default.
- W2122338824 date "2015-08-28" @default.
- W2122338824 modified "2023-10-16" @default.
- W2122338824 title "Vectorizing unstructured mesh computations for many‐core architectures" @default.
- W2122338824 cites W1988888548 @default.
- W2122338824 cites W1990062241 @default.
- W2122338824 cites W2005140610 @default.
- W2122338824 cites W2039985559 @default.
- W2122338824 cites W2053042479 @default.
- W2122338824 cites W2060803192 @default.
- W2122338824 cites W2095993144 @default.
- W2122338824 cites W2120385681 @default.
- W2122338824 cites W2138782497 @default.
- W2122338824 cites W2150947797 @default.
- W2122338824 cites W2160108531 @default.
- W2122338824 cites W2170996201 @default.
- W2122338824 cites W3010528205 @default.
- W2122338824 doi "https://doi.org/10.1002/cpe.3621" @default.
- W2122338824 hasPublicationYear "2015" @default.
- W2122338824 type Work @default.
- W2122338824 sameAs 2122338824 @default.
- W2122338824 citedByCount "14" @default.
- W2122338824 countsByYear W21223388242014 @default.
- W2122338824 countsByYear W21223388242015 @default.
- W2122338824 countsByYear W21223388242016 @default.
- W2122338824 countsByYear W21223388242017 @default.
- W2122338824 countsByYear W21223388242018 @default.
- W2122338824 countsByYear W21223388242019 @default.
- W2122338824 countsByYear W21223388242020 @default.
- W2122338824 crossrefType "journal-article" @default.
- W2122338824 hasAuthorship W2122338824A5035189605 @default.
- W2122338824 hasAuthorship W2122338824A5048478302 @default.
- W2122338824 hasAuthorship W2122338824A5056186758 @default.
- W2122338824 hasAuthorship W2122338824A5070789282 @default.
- W2122338824 hasBestOaLocation W21223388242 @default.
- W2122338824 hasConcept C111919701 @default.
- W2122338824 hasConcept C11413529 @default.
- W2122338824 hasConcept C13280743 @default.
- W2122338824 hasConcept C138101251 @default.
- W2122338824 hasConcept C145108525 @default.
- W2122338824 hasConcept C150552126 @default.
- W2122338824 hasConcept C173608175 @default.
- W2122338824 hasConcept C185798385 @default.
- W2122338824 hasConcept C205649164 @default.
- W2122338824 hasConcept C2778119891 @default.
- W2122338824 hasConcept C2908650547 @default.
- W2122338824 hasConcept C3826847 @default.
- W2122338824 hasConcept C41008148 @default.
- W2122338824 hasConcept C41681595 @default.
- W2122338824 hasConcept C45374587 @default.
- W2122338824 hasConcept C459310 @default.
- W2122338824 hasConcept C76752949 @default.
- W2122338824 hasConcept C78766204 @default.
- W2122338824 hasConcept C83283714 @default.
- W2122338824 hasConcept C96972482 @default.
- W2122338824 hasConceptScore W2122338824C111919701 @default.
- W2122338824 hasConceptScore W2122338824C11413529 @default.
- W2122338824 hasConceptScore W2122338824C13280743 @default.
- W2122338824 hasConceptScore W2122338824C138101251 @default.
- W2122338824 hasConceptScore W2122338824C145108525 @default.
- W2122338824 hasConceptScore W2122338824C150552126 @default.
- W2122338824 hasConceptScore W2122338824C173608175 @default.
- W2122338824 hasConceptScore W2122338824C185798385 @default.
- W2122338824 hasConceptScore W2122338824C205649164 @default.
- W2122338824 hasConceptScore W2122338824C2778119891 @default.
- W2122338824 hasConceptScore W2122338824C2908650547 @default.
- W2122338824 hasConceptScore W2122338824C3826847 @default.
- W2122338824 hasConceptScore W2122338824C41008148 @default.
- W2122338824 hasConceptScore W2122338824C41681595 @default.
- W2122338824 hasConceptScore W2122338824C45374587 @default.
- W2122338824 hasConceptScore W2122338824C459310 @default.
- W2122338824 hasConceptScore W2122338824C76752949 @default.
- W2122338824 hasConceptScore W2122338824C78766204 @default.
- W2122338824 hasConceptScore W2122338824C83283714 @default.
- W2122338824 hasConceptScore W2122338824C96972482 @default.
- W2122338824 hasFunder F4320320098 @default.
- W2122338824 hasIssue "2" @default.
- W2122338824 hasLocation W21223388241 @default.
- W2122338824 hasLocation W21223388242 @default.
- W2122338824 hasOpenAccess W2122338824 @default.
- W2122338824 hasPrimaryLocation W21223388241 @default.
- W2122338824 hasRelatedWork W1985658314 @default.
- W2122338824 hasRelatedWork W2020484966 @default.
- W2122338824 hasRelatedWork W2097822204 @default.
- W2122338824 hasRelatedWork W2122338824 @default.
- W2122338824 hasRelatedWork W2170268965 @default.
- W2122338824 hasRelatedWork W2526069705 @default.
- W2122338824 hasRelatedWork W2612377115 @default.
- W2122338824 hasRelatedWork W2613115449 @default.
- W2122338824 hasRelatedWork W2947212999 @default.
- W2122338824 hasRelatedWork W3083782334 @default.
- W2122338824 hasVolume "28" @default.
- W2122338824 isParatext "false" @default.