ETRI-Knowledge Sharing Plaform

KOREAN
논문 검색
Type SCI
Year ~ Keyword

Detail

Conference Paper Optically Networked Heterogeneous Data-centric Computing System with Silicon Photonics Transceivers
Cited - time in scopus Share share facebook twitter linkedin kakaostory
Authors
Dae-Ub Kim, Jyung Chan Lee, Sanghwa Yoo, Jongtae Song, Kyeong-Eun Han, Jiwook Youn, Bup Joong Kim, Chanho Park, Joon Ki Lee
Issue Date
2024-03
Citation
Optical Fiber Communication Conference (OFC) 2024, pp.1-3
Publisher
IEEE
Language
English
Type
Conference Paper
Abstract
An architecture is proposed for optically networked heterogeneous computing system supporting CXL standard. The proposed system is applied to AI applications and achieves a performance degradation of less than 2% compared with the server solution.
KSP Keywords
AI Applications, Data-centric, Heterogeneous computing systems, Heterogeneous data, Silicon photonics, performance degradation