AI and Memory Wall

Author:

Gholami Amir1ORCID,Yao Zhewei2ORCID,Kim Sehoon1ORCID,Hooper Coleman1ORCID,Mahoney Michael W.1ORCID,Keutzer Kurt1ORCID

Affiliation:

1. UC Berkeley, , Berkeley, CA, 94720, USA

2. Snowflake, Bellevue, , WA, 98004, USA

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Reference12 articles.

1. Language models are few-shot learners;Brown,2020

2. BERT: Pre-training of deep bidirectional transformers for language understanding;Devlin,2018

3. A Survey of Quantization Methods for Efficient Neural Network Inference

4. Sparsity in deep learning: Pruning and growth for efficient inference and training in neural networks;Hoefler;J. Mach. Learn. Res.,2021

5. Training compute-optimal large language models;Hoffmann,2022

Cited by 7 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

1. Efficient Inter-Datacenter AllReduce With Multiple Trees;IEEE Transactions on Network Science and Engineering;2024-09

2. Exploring the Viability of Composable Architectures to Overcome Memory Limitations in High Performance Computing Workflows;Practice and Experience in Advanced Research Computing 2024: Human Powered Computing;2024-07-17

3. An Automated Machine Learning Framework for Adaptive and Optimized Hyperspectral-Based Land Cover and Land-Use Segmentation;Remote Sensing;2024-07-12

4. Present and Future, Challenges of High Bandwith Memory (HBM);2024 IEEE International Memory Workshop (IMW);2024-05-12

5. An Incoming World of Decoupling Siliconomy;IEEE Micro;2024-05

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3