FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2891285910> ?p ?o ?g. }

• W2891285910 abstract "Graphics processing units (GPUs) feature an increasing number of streaming multiprocessors (SMs) with each successive generation. At the same time, GPUs are increasingly widely adopted in cloud services and data centers to accelerate general-purpose workloads. Running multiple applications on a GPU in such environments requires effective multitasking support. Spatial multitasking in which independent applications co-execute on different sets of SMs is a promising solution to share GPU resources. Unfortunately, how to effectively partition SMs is an open problem. In this paper, we observe that compared to widely-used even partitioning, dynamic SM partitioning based on the characteristics of the co-executing applications can significantly improve performance and power efficiency. Unfortunately finding an effective SM partition is challenging because the number of possible combinations increases exponentially with the number of SMs and co-executing applications. Through offline analysis, we find that first classifying workloads, and then searching an effective SM partition based on the workload characteristics can significantly reduce the search space, making dynamic SM partitioning tractable. Based on these insights, we propose Classification-Driven search (CD-search) for low-overhead dynamic SM partitioning in multitasking GPUs. CD-search first classifies workloads using a novel off-SM bandwidth model, after which it enters the performance mode or power mode depending on the workload's characteristics. Both modes follow a specific search strategy to quickly determine the optimum SM partition. Our evaluation shows that CD-search improves system throughput by 10.4% on average (and up to 62.9%) over even partitioning for workloads that are classified for the performance mode. For workloads classified for the power mode, CD-search reduces power consumption by 25% on average (and up to 41.2%). CD-search incurs limited runtime overhead." @default.
• W2891285910 created "2018-09-27" @default.
• W2891285910 creator A5033119975 @default.
• W2891285910 creator A5047295889 @default.
• W2891285910 creator A5072444630 @default.
• W2891285910 date "2018-06-12" @default.
• W2891285910 modified "2023-09-24" @default.
• W2891285910 title "Classification-Driven Search for Effective SM Partitioning in Multitasking GPUs" @default.
• W2891285910 cites W1973573211 @default.
• W2891285910 cites W2011389593 @default.
• W2891285910 cites W2080592089 @default.
• W2891285910 cites W2093043622 @default.
• W2891285910 cites W2100011668 @default.
• W2891285910 cites W2125551452 @default.
• W2891285910 cites W2129381159 @default.
• W2891285910 cites W2129817042 @default.
• W2891285910 cites W2150851481 @default.
• W2891285910 cites W2238992335 @default.
• W2891285910 cites W2289256051 @default.
• W2891285910 cites W2323909431 @default.
• W2891285910 cites W2605178034 @default.
• W2891285910 cites W2764065518 @default.
• W2891285910 cites W4243035950 @default.
• W2891285910 doi "https://doi.org/10.1145/3205289.3205311" @default.
• W2891285910 hasPublicationYear "2018" @default.
• W2891285910 type Work @default.
• W2891285910 sameAs 2891285910 @default.
• W2891285910 citedByCount "21" @default.
• W2891285910 countsByYear W28912859102019 @default.
• W2891285910 countsByYear W28912859102020 @default.
• W2891285910 countsByYear W28912859102021 @default.
• W2891285910 countsByYear W28912859102022 @default.
• W2891285910 countsByYear W28912859102023 @default.
• W2891285910 crossrefType "proceedings-article" @default.
• W2891285910 hasAuthorship W2891285910A5033119975 @default.
• W2891285910 hasAuthorship W2891285910A5047295889 @default.
• W2891285910 hasAuthorship W2891285910A5072444630 @default.
• W2891285910 hasConcept C107418235 @default.
• W2891285910 hasConcept C15744967 @default.
• W2891285910 hasConcept C173608175 @default.
• W2891285910 hasConcept C180747234 @default.
• W2891285910 hasConcept C41008148 @default.
• W2891285910 hasConceptScore W2891285910C107418235 @default.
• W2891285910 hasConceptScore W2891285910C15744967 @default.
• W2891285910 hasConceptScore W2891285910C173608175 @default.
• W2891285910 hasConceptScore W2891285910C180747234 @default.
• W2891285910 hasConceptScore W2891285910C41008148 @default.
• W2891285910 hasFunder F4320321001 @default.
• W2891285910 hasFunder F4320321730 @default.
• W2891285910 hasFunder F4320338335 @default.
• W2891285910 hasLocation W28912859101 @default.
• W2891285910 hasOpenAccess W2891285910 @default.
• W2891285910 hasPrimaryLocation W28912859101 @default.
• W2891285910 hasRelatedWork W1491899005 @default.
• W2891285910 hasRelatedWork W1604898313 @default.
• W2891285910 hasRelatedWork W2061998002 @default.
• W2891285910 hasRelatedWork W2117014006 @default.
• W2891285910 hasRelatedWork W2372170743 @default.
• W2891285910 hasRelatedWork W2888918673 @default.
• W2891285910 hasRelatedWork W2913044749 @default.
• W2891285910 hasRelatedWork W343461076 @default.
• W2891285910 hasRelatedWork W4231888100 @default.
• W2891285910 hasRelatedWork W4233815414 @default.
• W2891285910 isParatext "false" @default.
• W2891285910 isRetracted "false" @default.
• W2891285910 magId "2891285910" @default.
• W2891285910 workType "article" @default.