Zobrazeno 1 - 10
of 25
pro vyhledávání: '"Hongchan Roh"'
Autor:
Seokhyun Ryu, Sohyun Kim, Jaeyung Jun, Donguk Moon, Kyungsoo Lee, Jungmin Choi, Sunwoong Kim, Hyungsoo Kim, Luke Kim, Won Ha Choi, Moohyeon Nam, Dooyoung Hwang, Hongchan Roh, Youngpyo Joo
Publikováno v:
2023 IEEE International Conference on Big Data and Smart Computing (BigComp).
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::4b1801de03cfa3ae1c2f9cdba7ed9178
https://doi.org/10.1109/bigcomp57234.2023.00011
https://doi.org/10.1109/bigcomp57234.2023.00011
Publikováno v:
2022 IEEE 40th International Conference on Computer Design (ICCD).
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::69975d9e0dba83c2e5f5dbca05722bf5
https://doi.org/10.1109/iccd56317.2022.00051
https://doi.org/10.1109/iccd56317.2022.00051
Publikováno v:
Cluster Computing. 23:2287-2300
This paper presents a novel “Distributed Deep Learning Framework” for a heterogeneous multi-GPU cluster that can effectively improve overall resource utilization without sacrificing training accuracy. Specifically, we employ a hybrid aggregation
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::27b9efc745dbb6d112da0c99e2d6770a
https://doi.org/10.1007/s10586-020-03144-9
https://doi.org/10.1007/s10586-020-03144-9
Publikováno v:
KIISE Transactions on Computing Practices. 26:161-166
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::1bd33ce33533c4320c382adf61a34573
https://doi.org/10.5626/ktcp.2020.26.3.161
https://doi.org/10.5626/ktcp.2020.26.3.161
Publikováno v:
Cluster Computing. 23:2193-2204
This paper presents a comprehensive suite of techniques for optimized memory management in multi-GPU systems to accelerate deep learning application execution. We employ a hybrid utilization of GPU and CPU memories in a multi-GPU environment by effec
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::3d4d28e0ad63a8b672c7177f64fb1687
https://doi.org/10.1007/s10586-019-02974-6
https://doi.org/10.1007/s10586-019-02974-6
Publikováno v:
SAC
Recently, several studies show the powerful capability of neural networks to capture non-linear features from time series which have multiple seasonal patterns. However, existing methods rely on convolution kernels implicitly, hence neglect to captur
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::dc1a54223c2b845dd47300a9fd2342f1
https://doi.org/10.1145/3412841.3442108
https://doi.org/10.1145/3412841.3442108
Publikováno v:
IEEE Transactions on Computers. 67:589-595
Fast network quality analysis in the telecom industry is an important method used to provide quality service. SK Telecom, based in South Korea, built a Hadoop-based analytical system consisting of a hundred nodes, each of which only contains hard dis
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::3b4ca474d81b58907fb15773b9cb9d3e
https://doi.org/10.1109/tc.2017.2771491
https://doi.org/10.1109/tc.2017.2771491
Publikováno v:
IEEE Transactions on Knowledge and Data Engineering. 30:87-100
In this paper, we propose MV-FTL, a multi-version flash transition layer (FTL) that provides page-level multi-version management. By extending a unique characteristic of solid-state drives (SSDs), the out-of-place (OoP) update to multi-version manage
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::68da8e45b876d9a6329107a4da3483f2
https://doi.org/10.1109/tkde.2017.2757016
https://doi.org/10.1109/tkde.2017.2757016
Publikováno v:
FAS*W@SASO/ICAC
This paper presents a novel "Distributed Deep Learning Framework" for heterogeneous multi-GPU cluster that can effectively improve overall resource utilization without sacrificing training accuracy. Specifically, we employ a hybrid aggregation approa
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::19a2496275bc51657f1157bd3663e67e
https://doi.org/10.1109/fas-w.2019.00050
https://doi.org/10.1109/fas-w.2019.00050
Publikováno v:
BigComp
The in-memory key-value store provides a persistence method to ensure data durability. The currently provided methods are either to create a snapshot file of a current dataset or to write the log of the performed command in the log file. However, the
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::208ed3d9dcf05d7eeb406cefd42016e4
https://doi.org/10.1109/bigcomp.2019.8679377
https://doi.org/10.1109/bigcomp.2019.8679377