Matches in SemOpenAlex for { <https://semopenalex.org/work/W3136358373> ?p ?o ?g. }
Showing items 1 to 87 of
87
with 100 items per page.
- W3136358373 abstract "Developing efficient GPU kernels can be difficult because of the complexity of GPU architectures and programming models. Existing performance tools only provide coarse-grained tuning advice at the kernel level, if any. In this paper, we describe GPA, a performance advisor for NVIDIA GPUs that suggests potential code optimizations at a hierarchy of levels, including individual lines, loops, and functions. To relieve users of the burden of interpreting performance counters and analyzing bottlenecks, GPA uses data flow analysis to approximately attribute measured instruction stalls to their root causes and uses information about a program's structure and the GPU to match inefficiency patterns with optimization strategies. To quantify the potential benefits of each optimization strategy, we developed PC sampling-based performance models to estimate its speedup. Our experiments with benchmarks and applications show that GPA provides insightful reports to guide performance optimization. Using GPA, we obtained speedups on a Volta V100 GPU ranging from 1.01 x to 3.58 ×, with a geometric mean of 1.22 x." @default.
- W3136358373 created "2021-03-29" @default.
- W3136358373 creator A5027198310 @default.
- W3136358373 creator A5047472245 @default.
- W3136358373 creator A5063326523 @default.
- W3136358373 creator A5089709469 @default.
- W3136358373 date "2021-02-27" @default.
- W3136358373 modified "2023-09-25" @default.
- W3136358373 title "GPA: A GPU Performance Advisor Based on Instruction Sampling" @default.
- W3136358373 cites W1507654557 @default.
- W3136358373 cites W1517652255 @default.
- W3136358373 cites W1902930330 @default.
- W3136358373 cites W2013062050 @default.
- W3136358373 cites W2043218878 @default.
- W3136358373 cites W2080592089 @default.
- W3136358373 cites W2093226410 @default.
- W3136358373 cites W2136434791 @default.
- W3136358373 cites W2142079700 @default.
- W3136358373 cites W2290349115 @default.
- W3136358373 cites W2580538010 @default.
- W3136358373 cites W2767346422 @default.
- W3136358373 cites W2789572737 @default.
- W3136358373 cites W2952416601 @default.
- W3136358373 cites W2979340153 @default.
- W3136358373 cites W2989165654 @default.
- W3136358373 cites W2999812057 @default.
- W3136358373 cites W3015338905 @default.
- W3136358373 cites W3040626038 @default.
- W3136358373 cites W3132094421 @default.
- W3136358373 cites W4245644062 @default.
- W3136358373 cites W4248267595 @default.
- W3136358373 cites W4249818463 @default.
- W3136358373 cites W960901134 @default.
- W3136358373 doi "https://doi.org/10.1109/cgo51591.2021.9370339" @default.
- W3136358373 hasPublicationYear "2021" @default.
- W3136358373 type Work @default.
- W3136358373 sameAs 3136358373 @default.
- W3136358373 citedByCount "3" @default.
- W3136358373 countsByYear W31363583732021 @default.
- W3136358373 countsByYear W31363583732022 @default.
- W3136358373 countsByYear W31363583732023 @default.
- W3136358373 crossrefType "proceedings-article" @default.
- W3136358373 hasAuthorship W3136358373A5027198310 @default.
- W3136358373 hasAuthorship W3136358373A5047472245 @default.
- W3136358373 hasAuthorship W3136358373A5063326523 @default.
- W3136358373 hasAuthorship W3136358373A5089709469 @default.
- W3136358373 hasBestOaLocation W31363583732 @default.
- W3136358373 hasConcept C106131492 @default.
- W3136358373 hasConcept C118524514 @default.
- W3136358373 hasConcept C121684516 @default.
- W3136358373 hasConcept C140779682 @default.
- W3136358373 hasConcept C173608175 @default.
- W3136358373 hasConcept C21442007 @default.
- W3136358373 hasConcept C31972630 @default.
- W3136358373 hasConcept C41008148 @default.
- W3136358373 hasConcept C459310 @default.
- W3136358373 hasConcept C50630238 @default.
- W3136358373 hasConcept C86111242 @default.
- W3136358373 hasConceptScore W3136358373C106131492 @default.
- W3136358373 hasConceptScore W3136358373C118524514 @default.
- W3136358373 hasConceptScore W3136358373C121684516 @default.
- W3136358373 hasConceptScore W3136358373C140779682 @default.
- W3136358373 hasConceptScore W3136358373C173608175 @default.
- W3136358373 hasConceptScore W3136358373C21442007 @default.
- W3136358373 hasConceptScore W3136358373C31972630 @default.
- W3136358373 hasConceptScore W3136358373C41008148 @default.
- W3136358373 hasConceptScore W3136358373C459310 @default.
- W3136358373 hasConceptScore W3136358373C50630238 @default.
- W3136358373 hasConceptScore W3136358373C86111242 @default.
- W3136358373 hasLocation W31363583731 @default.
- W3136358373 hasLocation W31363583732 @default.
- W3136358373 hasOpenAccess W3136358373 @default.
- W3136358373 hasPrimaryLocation W31363583731 @default.
- W3136358373 hasRelatedWork W189420351 @default.
- W3136358373 hasRelatedWork W1978663335 @default.
- W3136358373 hasRelatedWork W2089453325 @default.
- W3136358373 hasRelatedWork W2128283661 @default.
- W3136358373 hasRelatedWork W2144511445 @default.
- W3136358373 hasRelatedWork W2176857686 @default.
- W3136358373 hasRelatedWork W2535207728 @default.
- W3136358373 hasRelatedWork W2570612679 @default.
- W3136358373 hasRelatedWork W2794923745 @default.
- W3136358373 hasRelatedWork W67367039 @default.
- W3136358373 isParatext "false" @default.
- W3136358373 isRetracted "false" @default.
- W3136358373 magId "3136358373" @default.
- W3136358373 workType "article" @default.