Matches in SemOpenAlex for { <https://semopenalex.org/work/W2943671750> ?p ?o ?g. }
Showing items 1 to 88 of
88
with 100 items per page.
- W2943671750 endingPage "24" @default.
- W2943671750 startingPage "1" @default.
- W2943671750 abstract "Graphics processing units (GPUs) 1 have enjoyed increasing popularity in recent years, which benefits from, for example, general-purpose GPU (GPGPU) for parallel programs and new computing paradigms, such as the Internet of Things (IoT). GPUs hold great potential in providing effective solutions for big data analytics while the demands for processing large quantities of data in real time are also increasing. However, the pervasive presence of GPUs on mobile devices presents great challenges for GPGPU, mainly because GPGPU integrates a large amount of processor arrays and concurrent executing threads (up to hundreds of thousands). In particular, the root causes of performance loss in a GPGPU program can not be revealed in detail by current approaches. In this article, we propose MiC (Multi-level Characterization), a framework that comprehensively characterizes GPGPU kernels at the instruction, Basic Block (BBL), and thread levels. Specifically, we devise Instruction Vectors (IV) and Basic Blocks Vectors (BBV), a Thread Similarity Matrix (TSM), and a Divergence Flow Statistics Graph (DFSG) to profile information in each level. We use MiC to provide insights into GPGPU kernels through the characterizations of 34 kernels from popular GPGPU benchmark suites such as Compute Unified Device Architecture (CUDA) Software Development Kit (SDK), Rodinia, and Parboil. In comparison with Central Processing Unit (CPU) workloads, we conclude the key findings as follows: (1) There are comparable Instruction-Level Parallelism (ILP); (2) The BBL count is significantly smaller than CPU workloads—only 22.8 on average; (3) The dynamic instruction count per thread varies from dozens to tens of thousands and it is extremely small compared to CPU benchmarks; (4) The Pareto principle (also called 90/10 rule) does not apply to GPGPU kernels while it pervasively exists in CPU programs; (5) The loop patterns are dramatically different from those in CPU workloads; (6) The branch ratio is lower than that of CPU programs but higher than pure GPU workloads. In addition, we have also shown how TSM and DFSG are used to characterize the branch divergence in a visual way, to enable the analysis of thread behavior in GPGPU programs. In addition, we show an optimization case for a GPGPU kernel from the bottleneck identified through its characterization result, which improves 16.8% performance." @default.
- W2943671750 created "2019-05-09" @default.
- W2943671750 creator A5048614443 @default.
- W2943671750 creator A5056830118 @default.
- W2943671750 creator A5069348195 @default.
- W2943671750 date "2019-04-29" @default.
- W2943671750 modified "2023-09-24" @default.
- W2943671750 title "MiC" @default.
- W2943671750 cites W1971997351 @default.
- W2943671750 cites W1992851788 @default.
- W2943671750 cites W2080592089 @default.
- W2943671750 cites W2090584832 @default.
- W2943671750 cites W2098505406 @default.
- W2943671750 cites W2099135670 @default.
- W2943671750 cites W2099891946 @default.
- W2943671750 cites W2101366948 @default.
- W2943671750 cites W2101587002 @default.
- W2943671750 cites W2106018915 @default.
- W2943671750 cites W2107493680 @default.
- W2943671750 cites W2128022558 @default.
- W2943671750 cites W2145913625 @default.
- W2943671750 cites W2149693551 @default.
- W2943671750 cites W2153456949 @default.
- W2943671750 cites W2155503253 @default.
- W2943671750 cites W2155568054 @default.
- W2943671750 cites W2156831150 @default.
- W2943671750 cites W2158924248 @default.
- W2943671750 cites W2169880332 @default.
- W2943671750 cites W2236252626 @default.
- W2943671750 cites W2344044374 @default.
- W2943671750 cites W2517714857 @default.
- W2943671750 cites W2528187657 @default.
- W2943671750 cites W2730940305 @default.
- W2943671750 doi "https://doi.org/10.1145/3304108" @default.
- W2943671750 hasPublicationYear "2019" @default.
- W2943671750 type Work @default.
- W2943671750 sameAs 2943671750 @default.
- W2943671750 citedByCount "1" @default.
- W2943671750 countsByYear W29436717502022 @default.
- W2943671750 crossrefType "journal-article" @default.
- W2943671750 hasAuthorship W2943671750A5048614443 @default.
- W2943671750 hasAuthorship W2943671750A5056830118 @default.
- W2943671750 hasAuthorship W2943671750A5069348195 @default.
- W2943671750 hasConcept C111919701 @default.
- W2943671750 hasConcept C13280743 @default.
- W2943671750 hasConcept C138101251 @default.
- W2943671750 hasConcept C173608175 @default.
- W2943671750 hasConcept C185798385 @default.
- W2943671750 hasConcept C205649164 @default.
- W2943671750 hasConcept C21442007 @default.
- W2943671750 hasConcept C2778119891 @default.
- W2943671750 hasConcept C2779851693 @default.
- W2943671750 hasConcept C41008148 @default.
- W2943671750 hasConcept C50630238 @default.
- W2943671750 hasConceptScore W2943671750C111919701 @default.
- W2943671750 hasConceptScore W2943671750C13280743 @default.
- W2943671750 hasConceptScore W2943671750C138101251 @default.
- W2943671750 hasConceptScore W2943671750C173608175 @default.
- W2943671750 hasConceptScore W2943671750C185798385 @default.
- W2943671750 hasConceptScore W2943671750C205649164 @default.
- W2943671750 hasConceptScore W2943671750C21442007 @default.
- W2943671750 hasConceptScore W2943671750C2778119891 @default.
- W2943671750 hasConceptScore W2943671750C2779851693 @default.
- W2943671750 hasConceptScore W2943671750C41008148 @default.
- W2943671750 hasConceptScore W2943671750C50630238 @default.
- W2943671750 hasFunder F4320321001 @default.
- W2943671750 hasFunder F4320321543 @default.
- W2943671750 hasIssue "3" @default.
- W2943671750 hasLocation W29436717501 @default.
- W2943671750 hasOpenAccess W2943671750 @default.
- W2943671750 hasPrimaryLocation W29436717501 @default.
- W2943671750 hasRelatedWork W1790810173 @default.
- W2943671750 hasRelatedWork W1994718705 @default.
- W2943671750 hasRelatedWork W2075046026 @default.
- W2943671750 hasRelatedWork W2085124982 @default.
- W2943671750 hasRelatedWork W2320652536 @default.
- W2943671750 hasRelatedWork W2401424361 @default.
- W2943671750 hasRelatedWork W2770405949 @default.
- W2943671750 hasRelatedWork W2794923745 @default.
- W2943671750 hasRelatedWork W564427 @default.
- W2943671750 hasRelatedWork W2945926249 @default.
- W2943671750 hasVolume "15" @default.
- W2943671750 isParatext "false" @default.
- W2943671750 isRetracted "false" @default.
- W2943671750 magId "2943671750" @default.
- W2943671750 workType "article" @default.