Zobrazeno 1 - 10
of 28
pro vyhledávání: '"Hyojin Sung"'
Autor:
Hyungkyu Ham, Hyunuk Cho, Minjae Kim, Jueon Park, Jeongmin Hong, Hyojin Sung, Eunhyeok Park, Euicheol Lim, Gwangsun Kim
Publikováno v:
IEEE Access, Vol 12, Pp 142651-142667 (2024)
Currently, GPUs face significant challenges due to limited off-chip bandwidth (BW) and memory capacity during DNN training. To address these bottlenecks, we propose a memory access-triggered near-data processing matNDP architecture that offloads memo
Externí odkaz:
https://doaj.org/article/5e0733ab887241b48181918dc53d2a04
Autor:
Jueon Park, Hyojin Sung
Publikováno v:
IEEE Computer Architecture Letters. :1-4
Publikováno v:
IEEE Computer Architecture Letters. 21:33-36
Publikováno v:
Proceedings of the 21st ACM/IEEE International Symposium on Code Generation and Optimization.
Artifact for PIMFlow. For development environment setup, refer to Docker Hub. The latest codes are on GitHub. To skip trace generation, unzip data.zip and move all *.tar.gz files to PIMFlow/data.
Artifact evaluated final version
Artifact evaluated final version
Artifact for PIMFlow. For development environment setup, refer toDocker Hub. The latest codes are onGitHub. To skip trace generation, unzip data.zip andmove all *.tar.gz filesto PIMFlow/data.
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::499b25fae53ee132f72de14a02708c6b
Autor:
Eunhyeok Park, Jeongmin Hong, Euicheol Lim, Hyunuk Cho, Minjae Kim, Gwangsun Kim, Jueon Park, Hyungkyu Ham, Hyojin Sung
Publikováno v:
IEEE Computer Architecture Letters. 20:171-174
Publikováno v:
Proceedings of the 31st ACM SIGPLAN International Conference on Compiler Construction.
Autor:
Yongwon Shin, Hyojin Sung
Publikováno v:
Languages and Compilers for Parallel Computing ISBN: 9783030993719
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::42d3fa21b91e5e31a9cff37703ae2096
https://doi.org/10.1007/978-3-030-99372-6_3
https://doi.org/10.1007/978-3-030-99372-6_3
Publikováno v:
Journal of KIISE. 45:99-105
Publikováno v:
PACT
We propose CogR, a machine-learning based runtime solution, that enables efficient and dynamic resource scheduling and performance optimization for high-level programming interfaces on heterogeneous systems. CogR tightly combines the structural infor