SpecPIM: Accelerating Speculative Inference on PIM-Enabled System via Architecture-Dataflow Co-Exploration-Reference-Cited by-同舟云学术

登录注册会员服务联系我们

SpecPIM: Accelerating Speculative Inference on PIM-Enabled System via Architecture-Dataflow Co-Exploration

Published:2024-04-27 Issue: Volume:1 Page:950-965
ISSN:
Container-title:Proceedings of the 29th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 3
language:
Short-container-title:

Author:

Li Cong¹^ORCID,Zhou Zhe²^ORCID,Zheng Size²^ORCID,Zhang Jiaxi³^ORCID,Liang Yun¹⁴^ORCID,Sun Guangyu¹⁴^ORCID

Affiliation:

1. School of Integrated Circuits, Peking University, Beijing, China

2. School of Integrated Circuits, School of Computer Science, Peking University, Beijing, China

3. School of Computer Science, Peking University, Beijing, China

4. Beijing Advanced Innovation Center for Integrated Circuits, Beijing, China

Funder

National Natural Science Foundation of China

111 Project

Natural Science Foundation of China

Publisher

ACM

Reference76 articles.

1. A scalable processing-in-memory accelerator for parallel graph processing

2. Amazon. Bedrock. https://aws.amazon.com/bedrock/.

3. DeepSpeed- Inference: Enabling Efficient Inference of Transformer Models at Unprecedented Scale

4. FAFNIR: Accelerating Sparse Gathering by Using Efficient Near-Memory Intelligent Reduction

5. Hadi Asghari-Moghaddam, Young Hoon Son, Jung Ho Ahn, and Nam Sung Kim. Chameleon: Versatile and practical near-dram acceleration architecture for large memory systems. In 2016 49th annual IEEE/ACM international symposium on Microarchitecture (MICRO), pages 1--13. IEEE, 2016.

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