Architecture Support for Domain-Specific Accelerator-Rich CMPs

Author:

Cong Jason1,Ghodrat Mohammad Ali1,Gill Michael1,Grigorian Beayna1,Reinman Glenn1

Affiliation:

1. University of California, Los Angeles

Abstract

This work discusses hardware architectural support for domain-specific accelerator-rich CMPs. First, we present a hardware resource management scheme for sharing of loosely coupled accelerators and arbitration of multiple requesting cores. Second, we present a mechanism for accelerator virtualization. This allows multiple accelerators to efficiently compose a larger virtual accelerator out of multiple smaller accelerators, as well as to collaborate as multiple copies of a simple accelerator. All of this work is supported by a fully automated simulation tool-chain for both accelerator generation and management. We present the applicability of our approach to four different application domains: medical imaging, commercial, computer vision, and navigation. Our results demonstrate large performance improvements and energy savings over a software implementation. We also show additional improvements that result from enhanced load balancing and simplification of the communication between the core and the arbitration mechanism.

Funder

C-FAR

Division of Graduate Education

Semiconductor Research Corporation

Division of Computing and Communication Foundations

Publisher

Association for Computing Machinery (ACM)

Subject

Hardware and Architecture,Software

Reference29 articles.

Cited by 16 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. RELIEF: Relieving Memory Pressure In SoCs Via Data Movement-Aware Accelerator Scheduling;2024 IEEE International Symposium on High-Performance Computer Architecture (HPCA);2024-03-02

2. Hardware Implementation of RF Spectral Convergence System on DASH SoC;2023 57th Asilomar Conference on Signals, Systems, and Computers;2023-10-29

3. The DASH SoC: Enabling the Next Generation of Multi-Function RF Systems;2022 56th Asilomar Conference on Signals, Systems, and Computers;2022-10-31

4. Technological Advances to Facilitate Spectral Convergence;2021 55th Asilomar Conference on Signals, Systems, and Computers;2021-10-31

5. Maximizing the Serviceability of Partially Reconfigurable FPGA Systems in Multi-tenant Environment;Proceedings of the 2020 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays;2020-02-23

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