Low-Precision Hardware Architectures Meet Recommendation Model Inference at Scale

Paper Info

Title
Low-Precision Hardware Architectures Meet Recommendation Model Inference at Scale

Abstract
Tremendous success of machine learning (ML) and the unabated growth in model complexity motivated many ML-specific designs in hardware architectures to speed up the model inference. While these architectures are diverse, highly optimized low-precision arithmetic is a component shared by most. Nevertheless, recommender systems important to Facebook’s personalization services are demanding and compl...

Year	DOI	Venue
2021	10.1109/MM.2021.3081981	IEEE Micro
Keywords	DocType	Volume
Quantization (signal),Computational modeling,Production,Computer architecture,Adaptation models,Predictive models,Training	Journal	41
Issue	ISSN	Citations
5	0272-1732	0
PageRank	References	Authors
0.34	0	20

Authors (20 rows)

Cited by (0 rows)

References (0 rows)

Name	Order	Citations	PageRank
Zhaoxia	1	0	0.34
Deng	2	0	0.34
Jongsoo Park	3	103	9.49
Ping Tak Peter Tang	4	229	12.50
Haixin Liu	5	0	1.35
Jie	6	0	0.34
Yang	7	41	8.06
Hector Yuen	8	6	1.10
Jianyu Huang	9	9	2.83
Daya Khudia	10	0	0.34
Xiaohan Wei	11	0	1.69
Ellie Wen	12	0	0.34
Dhruv Choudhary	13	61	6.60
Raghuraman Krishnamoorthi	14	26	1.85
Carole-Jean Wu	15	0	1.35
Nadathur Satish	16	2020	99.88
Changkyu Kim	17	1848	98.89
Maxim Naumov	18	0	0.34
Sam Naghshineh	19	0	0.34
Mikhail Smelyanskiy	20	4	1.06