Matches in SemOpenAlex for { <https://semopenalex.org/work/W2754674485> ?p ?o ?g. }
Showing items 1 to 98 of
98
with 100 items per page.
- W2754674485 endingPage "18974" @default.
- W2754674485 startingPage "18953" @default.
- W2754674485 abstract "Using FPGA-based acceleration of high-performance computing (HPC) applications to reduce energy and power consumption is becoming an interesting option, thanks to the availability of high-level synthesis (HLS) tools that enable fast design cycles. However, obtaining good performance for memoryintensive algorithms, which often exchange large data arrays with external DRAM, still requires timeconsuming optimization and good knowledge of hardware design. This article proposes a new design methodology, based on dedicated applicationand data array-specific caches. These caches provide most of the benefits that can be achieved by coding optimized DMA-like transfer strategies by hand into the HPC application code, but require only limited manual tuning (basically the selection of architecture and size), are neutral to target HLS tool and technology (FPGA or ASIC), and do not require changes to application code. We show experimental results obtained on five common memory-intensive algorithms from very diverse domains, namely machine learning, data sorting, and computer vision. We test the cost and performance of our caches against both out-of-the-box code originally optimized for a GPU, and manually optimized implementations specifically targeted for FPGAs via HLS. The implementation using our caches achieved an 8X speedup and 2X energy reduction on average with respect to out-of-the-box models using only simple directive-based optimizations (e.g., pipelining). They also achieved comparable performance with much less design effort when compared with the versions that were manually optimized to achieve efficient memory transfers specifically for an FPGA." @default.
- W2754674485 created "2017-09-25" @default.
- W2754674485 creator A5030646218 @default.
- W2754674485 creator A5038226992 @default.
- W2754674485 creator A5050365912 @default.
- W2754674485 creator A5065503853 @default.
- W2754674485 date "2017-01-01" @default.
- W2754674485 modified "2023-09-30" @default.
- W2754674485 title "Acceleration by Inline Cache for Memory-Intensive Algorithms on FPGA via High-Level Synthesis" @default.
- W2754674485 cites W2005022399 @default.
- W2754674485 cites W2013594929 @default.
- W2754674485 cites W2022128349 @default.
- W2754674485 cites W2061624656 @default.
- W2754674485 cites W2064115172 @default.
- W2754674485 cites W2087064593 @default.
- W2754674485 cites W2097537332 @default.
- W2754674485 cites W2109683087 @default.
- W2754674485 cites W2115069592 @default.
- W2754674485 cites W2118558147 @default.
- W2754674485 cites W2127788760 @default.
- W2754674485 cites W2138411892 @default.
- W2754674485 cites W2139774022 @default.
- W2754674485 cites W2149294210 @default.
- W2754674485 cites W2165949232 @default.
- W2754674485 cites W2286366494 @default.
- W2754674485 cites W2527148519 @default.
- W2754674485 cites W2560652762 @default.
- W2754674485 cites W2564844125 @default.
- W2754674485 cites W2589329959 @default.
- W2754674485 cites W3003468330 @default.
- W2754674485 cites W3145738314 @default.
- W2754674485 cites W4233785673 @default.
- W2754674485 cites W4237150160 @default.
- W2754674485 doi "https://doi.org/10.1109/access.2017.2750923" @default.
- W2754674485 hasPublicationYear "2017" @default.
- W2754674485 type Work @default.
- W2754674485 sameAs 2754674485 @default.
- W2754674485 citedByCount "7" @default.
- W2754674485 countsByYear W27546744852018 @default.
- W2754674485 countsByYear W27546744852019 @default.
- W2754674485 countsByYear W27546744852020 @default.
- W2754674485 countsByYear W27546744852021 @default.
- W2754674485 countsByYear W27546744852022 @default.
- W2754674485 countsByYear W27546744852023 @default.
- W2754674485 crossrefType "journal-article" @default.
- W2754674485 hasAuthorship W2754674485A5030646218 @default.
- W2754674485 hasAuthorship W2754674485A5038226992 @default.
- W2754674485 hasAuthorship W2754674485A5050365912 @default.
- W2754674485 hasAuthorship W2754674485A5065503853 @default.
- W2754674485 hasBestOaLocation W27546744851 @default.
- W2754674485 hasConcept C11413529 @default.
- W2754674485 hasConcept C115537543 @default.
- W2754674485 hasConcept C118524514 @default.
- W2754674485 hasConcept C119599485 @default.
- W2754674485 hasConcept C127413603 @default.
- W2754674485 hasConcept C149635348 @default.
- W2754674485 hasConcept C173608175 @default.
- W2754674485 hasConcept C2742236 @default.
- W2754674485 hasConcept C41008148 @default.
- W2754674485 hasConcept C42935608 @default.
- W2754674485 hasConcept C58013763 @default.
- W2754674485 hasConcept C68339613 @default.
- W2754674485 hasConcept C77390884 @default.
- W2754674485 hasConceptScore W2754674485C11413529 @default.
- W2754674485 hasConceptScore W2754674485C115537543 @default.
- W2754674485 hasConceptScore W2754674485C118524514 @default.
- W2754674485 hasConceptScore W2754674485C119599485 @default.
- W2754674485 hasConceptScore W2754674485C127413603 @default.
- W2754674485 hasConceptScore W2754674485C149635348 @default.
- W2754674485 hasConceptScore W2754674485C173608175 @default.
- W2754674485 hasConceptScore W2754674485C2742236 @default.
- W2754674485 hasConceptScore W2754674485C41008148 @default.
- W2754674485 hasConceptScore W2754674485C42935608 @default.
- W2754674485 hasConceptScore W2754674485C58013763 @default.
- W2754674485 hasConceptScore W2754674485C68339613 @default.
- W2754674485 hasConceptScore W2754674485C77390884 @default.
- W2754674485 hasFunder F4320320300 @default.
- W2754674485 hasLocation W27546744851 @default.
- W2754674485 hasLocation W27546744852 @default.
- W2754674485 hasOpenAccess W2754674485 @default.
- W2754674485 hasPrimaryLocation W27546744851 @default.
- W2754674485 hasRelatedWork W1567049613 @default.
- W2754674485 hasRelatedWork W1604320855 @default.
- W2754674485 hasRelatedWork W2062495483 @default.
- W2754674485 hasRelatedWork W2066442567 @default.
- W2754674485 hasRelatedWork W2100470915 @default.
- W2754674485 hasRelatedWork W2104020799 @default.
- W2754674485 hasRelatedWork W2383279219 @default.
- W2754674485 hasRelatedWork W2476941693 @default.
- W2754674485 hasRelatedWork W3010492628 @default.
- W2754674485 hasRelatedWork W2506672464 @default.
- W2754674485 hasVolume "5" @default.
- W2754674485 isParatext "false" @default.
- W2754674485 isRetracted "false" @default.
- W2754674485 magId "2754674485" @default.
- W2754674485 workType "article" @default.