Matches in SemOpenAlex for { <https://semopenalex.org/work/W2912779600> ?p ?o ?g. }
Showing items 1 to 58 of
58
with 100 items per page.
- W2912779600 abstract "GPU utilization, measured as occupancy, is limited by the parallel threads' combined usage of on-chip resources. If the resource demand cannot be met, GPUs will reduce the number of concurrent threads, impacting the program performance. We have observed that registers are the occupancy limiters while shared metmory tends to be underused. The de facto approach spills excessive registers to the out-of-chip memory, ignoring the shared memory and leaving the on-chip resources underutilized. To mitigate the register demand, our work presents a novel compiler technique, called register demotion, that allows data in the register to be placed into the underutilized shared memory by transforming the GPU assembly code (SASS). Register demotion achieves up to 18% speedup over the nvcc compiler, with a geometric mean of 7%." @default.
- W2912779600 created "2019-02-21" @default.
- W2912779600 creator A5000932612 @default.
- W2912779600 creator A5008393837 @default.
- W2912779600 creator A5045622261 @default.
- W2912779600 date "2019-02-16" @default.
- W2912779600 modified "2023-09-25" @default.
- W2912779600 title "Optimizing GPU programs by register demotion" @default.
- W2912779600 cites W1870686413 @default.
- W2912779600 cites W2078994750 @default.
- W2912779600 cites W2126830109 @default.
- W2912779600 cites W2149234156 @default.
- W2912779600 cites W2232645663 @default.
- W2912779600 cites W2408916298 @default.
- W2912779600 doi "https://doi.org/10.1145/3293883.3297859" @default.
- W2912779600 hasPublicationYear "2019" @default.
- W2912779600 type Work @default.
- W2912779600 sameAs 2912779600 @default.
- W2912779600 citedByCount "0" @default.
- W2912779600 crossrefType "proceedings-article" @default.
- W2912779600 hasAuthorship W2912779600A5000932612 @default.
- W2912779600 hasAuthorship W2912779600A5008393837 @default.
- W2912779600 hasAuthorship W2912779600A5045622261 @default.
- W2912779600 hasConcept C104545631 @default.
- W2912779600 hasConcept C138885662 @default.
- W2912779600 hasConcept C173608175 @default.
- W2912779600 hasConcept C17744445 @default.
- W2912779600 hasConcept C199539241 @default.
- W2912779600 hasConcept C2779235478 @default.
- W2912779600 hasConcept C41008148 @default.
- W2912779600 hasConcept C41895202 @default.
- W2912779600 hasConcept C94625758 @default.
- W2912779600 hasConceptScore W2912779600C104545631 @default.
- W2912779600 hasConceptScore W2912779600C138885662 @default.
- W2912779600 hasConceptScore W2912779600C173608175 @default.
- W2912779600 hasConceptScore W2912779600C17744445 @default.
- W2912779600 hasConceptScore W2912779600C199539241 @default.
- W2912779600 hasConceptScore W2912779600C2779235478 @default.
- W2912779600 hasConceptScore W2912779600C41008148 @default.
- W2912779600 hasConceptScore W2912779600C41895202 @default.
- W2912779600 hasConceptScore W2912779600C94625758 @default.
- W2912779600 hasLocation W29127796001 @default.
- W2912779600 hasOpenAccess W2912779600 @default.
- W2912779600 hasPrimaryLocation W29127796001 @default.
- W2912779600 hasRelatedWork W1491899005 @default.
- W2912779600 hasRelatedWork W1502414128 @default.
- W2912779600 hasRelatedWork W1558545464 @default.
- W2912779600 hasRelatedWork W1604898313 @default.
- W2912779600 hasRelatedWork W1984303163 @default.
- W2912779600 hasRelatedWork W2117014006 @default.
- W2912779600 hasRelatedWork W2172791042 @default.
- W2912779600 hasRelatedWork W2372170743 @default.
- W2912779600 hasRelatedWork W2790489068 @default.
- W2912779600 hasRelatedWork W4233815414 @default.
- W2912779600 isParatext "false" @default.
- W2912779600 isRetracted "false" @default.
- W2912779600 magId "2912779600" @default.
- W2912779600 workType "article" @default.