Title

SEMS: Scalable Embedding Memory System Exploiting Near-data Processing

DGIST Authors

Sejin Kim ; Sungjin Lee ; Jaeha Kung

Advisor

이성진

Co-Advisor(s)

Jaeha Kung

Issued Date

2022

Awarded Date

2022/02

Citation

Sejin Kim. (2022). SEMS: Scalable Embedding Memory System Exploiting Near-data Processing. doi: 10.22677/thesis.200000594560

Type

Thesis

Subject

Near-data processing, Machine learning, Embedding, System architecture

Description

Near-data processing, Machine learning, Embedding, System architecture

Table Of Contents

I. Introduction 1
II. Background 3
2.1 Embedding Layer 3
2.2 Recommender System 4
III. Motivation 6
IV. Design of SEMS 9
4.1 Overall Architecture 9
4.2 Near-data Processing 10
4.3 Developing Applications 12
4.4 Interface 12
4.5 Model Parallelism 15
4.6 Optimization 18
V. Evaluation 20
5.1 Methodology 20
5.2 Experimental Results 21
VI. Related Work 26
VII. Conclusion 28
References 29

URI

http://dgist.dcollection.net/common/orgView/200000594560
http://hdl.handle.net/20.500.11750/16324

DOI

10.22677/thesis.200000594560

Degree

Master

Department

Information and Communication Engineering

Publisher

DGIST

Show Full Item Record