Matches in SemOpenAlex for { <https://semopenalex.org/work/W4284975144> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W4284975144 abstract "Upcoming regional and National Science Foundation (NSF)-funded Cyberinfrastructure (CI) resources will give researchers opportunities to run their artificial intelligence / machine learning (AI/ML) workflows on accelerators. To effectively leverage this burgeoning CI-rich landscape, researchers need extensive benchmark data to maximize performance gains and map their workflows to appropriate architectures. This data will further assist CI administrators, NSF program officers, and CI allocation-reviewers make informed determinations on CI-resource allocations. Here, we compare the performance of two very different architectures: the commonly used Graphical Processing Units (GPUs) and the new generation of Intelligence Processing Units (IPUs), by running training benchmarks of common AI/ML models. We leverage the maturity of software stacks, and the ease of migration among these platforms to learn that performance and scaling are similar for both architectures. Exploring training parameters, such as batch size, however finds that owing to memory processing structures, IPUs run efficiently with smaller batch sizes, while GPUs benefit from large batch sizes to extract sufficient parallelism in neural network training and inference. This comes with different advantages and disadvantages as discussed in this paper.As such considerations of inference latency, inherent parallelism and model accuracy will play a role in researcher selection of these architectures. The impact of these choices on a representative image compression model system is discussed." @default.
- W4284975144 created "2022-07-10" @default.
- W4284975144 creator A5004379411 @default.
- W4284975144 creator A5013964237 @default.
- W4284975144 creator A5014565735 @default.
- W4284975144 creator A5017984059 @default.
- W4284975144 creator A5039844494 @default.
- W4284975144 creator A5043257503 @default.
- W4284975144 creator A5048746854 @default.
- W4284975144 creator A5063166697 @default.
- W4284975144 creator A5069831701 @default.
- W4284975144 creator A5070232707 @default.
- W4284975144 creator A5084608346 @default.
- W4284975144 creator A5090685598 @default.
- W4284975144 date "2022-07-08" @default.
- W4284975144 modified "2023-10-18" @default.
- W4284975144 title "Benchmarking the Performance of Accelerators on National Cyberinfrastructure Resources for Artificial Intelligence / Machine Learning Workloads" @default.
- W4284975144 cites W1997162567 @default.
- W4284975144 cites W2024585065 @default.
- W4284975144 cites W2071039340 @default.
- W4284975144 cites W2618530766 @default.
- W4284975144 doi "https://doi.org/10.1145/3491418.3530772" @default.
- W4284975144 hasPublicationYear "2022" @default.
- W4284975144 type Work @default.
- W4284975144 citedByCount "1" @default.
- W4284975144 countsByYear W42849751442023 @default.
- W4284975144 crossrefType "proceedings-article" @default.
- W4284975144 hasAuthorship W4284975144A5004379411 @default.
- W4284975144 hasAuthorship W4284975144A5013964237 @default.
- W4284975144 hasAuthorship W4284975144A5014565735 @default.
- W4284975144 hasAuthorship W4284975144A5017984059 @default.
- W4284975144 hasAuthorship W4284975144A5039844494 @default.
- W4284975144 hasAuthorship W4284975144A5043257503 @default.
- W4284975144 hasAuthorship W4284975144A5048746854 @default.
- W4284975144 hasAuthorship W4284975144A5063166697 @default.
- W4284975144 hasAuthorship W4284975144A5069831701 @default.
- W4284975144 hasAuthorship W4284975144A5070232707 @default.
- W4284975144 hasAuthorship W4284975144A5084608346 @default.
- W4284975144 hasAuthorship W4284975144A5090685598 @default.
- W4284975144 hasConcept C108583219 @default.
- W4284975144 hasConcept C119857082 @default.
- W4284975144 hasConcept C13280743 @default.
- W4284975144 hasConcept C144133560 @default.
- W4284975144 hasConcept C153083717 @default.
- W4284975144 hasConcept C154945302 @default.
- W4284975144 hasConcept C162853370 @default.
- W4284975144 hasConcept C173608175 @default.
- W4284975144 hasConcept C177212765 @default.
- W4284975144 hasConcept C185798385 @default.
- W4284975144 hasConcept C205649164 @default.
- W4284975144 hasConcept C2522767166 @default.
- W4284975144 hasConcept C2776214188 @default.
- W4284975144 hasConcept C2776397876 @default.
- W4284975144 hasConcept C2781172179 @default.
- W4284975144 hasConcept C41008148 @default.
- W4284975144 hasConcept C50644808 @default.
- W4284975144 hasConcept C61483411 @default.
- W4284975144 hasConcept C77088390 @default.
- W4284975144 hasConcept C86251818 @default.
- W4284975144 hasConceptScore W4284975144C108583219 @default.
- W4284975144 hasConceptScore W4284975144C119857082 @default.
- W4284975144 hasConceptScore W4284975144C13280743 @default.
- W4284975144 hasConceptScore W4284975144C144133560 @default.
- W4284975144 hasConceptScore W4284975144C153083717 @default.
- W4284975144 hasConceptScore W4284975144C154945302 @default.
- W4284975144 hasConceptScore W4284975144C162853370 @default.
- W4284975144 hasConceptScore W4284975144C173608175 @default.
- W4284975144 hasConceptScore W4284975144C177212765 @default.
- W4284975144 hasConceptScore W4284975144C185798385 @default.
- W4284975144 hasConceptScore W4284975144C205649164 @default.
- W4284975144 hasConceptScore W4284975144C2522767166 @default.
- W4284975144 hasConceptScore W4284975144C2776214188 @default.
- W4284975144 hasConceptScore W4284975144C2776397876 @default.
- W4284975144 hasConceptScore W4284975144C2781172179 @default.
- W4284975144 hasConceptScore W4284975144C41008148 @default.
- W4284975144 hasConceptScore W4284975144C50644808 @default.
- W4284975144 hasConceptScore W4284975144C61483411 @default.
- W4284975144 hasConceptScore W4284975144C77088390 @default.
- W4284975144 hasConceptScore W4284975144C86251818 @default.
- W4284975144 hasFunder F4320306076 @default.
- W4284975144 hasLocation W42849751441 @default.
- W4284975144 hasOpenAccess W4284975144 @default.
- W4284975144 hasPrimaryLocation W42849751441 @default.
- W4284975144 hasRelatedWork W2148010638 @default.
- W4284975144 hasRelatedWork W2428494321 @default.
- W4284975144 hasRelatedWork W2442273404 @default.
- W4284975144 hasRelatedWork W2593649365 @default.
- W4284975144 hasRelatedWork W2950577464 @default.
- W4284975144 hasRelatedWork W3098379356 @default.
- W4284975144 hasRelatedWork W3214188003 @default.
- W4284975144 hasRelatedWork W3214520715 @default.
- W4284975144 hasRelatedWork W4281488864 @default.
- W4284975144 hasRelatedWork W4226055497 @default.
- W4284975144 isParatext "false" @default.
- W4284975144 isRetracted "false" @default.
- W4284975144 workType "article" @default.