一种基于数据流解耦的可重构阵列映射方法

钱恺宸; 尹琛; 景乃锋

Chinese

您当前的位置：

首页 >

文章列表页 >

一种基于数据流解耦的可重构阵列映射方法

更新时间：2026-09-04

- 一种基于数据流解耦的可重构阵列映射方法
- A mapping method for reconfigurable array based on data flow decoupling
- 微电子学与计算机 2021年38卷第8期页码：28-33
- 作者机构：
  
  上海交通大学电子信息与电气工程学院,上海,200240
- 作者简介：
- 基金信息：
- DOI：
  中图分类号：
- 纸质出版：2021
- 稿件说明：
移动端阅览
钱恺宸, 尹琛, 景乃锋. 一种基于数据流解耦的可重构阵列映射方法[J]. 微电子学与计算机, 2021, 38(8): 28-33.

QIAN Kaichen, YIN Chen, JING Naifeng. A mapping method for reconfigurable array based on data flow decoupling[J]. 2021, 38(8): 28-33.
钱恺宸, 尹琛, 景乃锋. 一种基于数据流解耦的可重构阵列映射方法[J]. 微电子学与计算机, 2021, 38(8): 28-33. DOI：

QIAN Kaichen, YIN Chen, JING Naifeng. A mapping method for reconfigurable array based on data flow decoupling[J]. 2021, 38(8): 28-33. DOI：

摘要

可重构阵列依靠数据流驱动带来的能效优势被作为加速器广泛运用在特定领域之中.随着应用范围的增大，当应用中存在不同执行速率的区域时，采用传统的空间映射方案将整个数据流图进行直接映射会造成严重的性能损失.提出了一种基于数据流解耦的映射方法，通过在执行速率不同的区域之间加入解耦单元以解决执行速率失配的问题.同时利用一种分布式多阶段的布局算法，将解耦后的数据流图映射在分簇式的互连结构中.与传统映射方案相比，对数据流进行解耦后的映射方案在保持较高的互连布通率的基础上可以平均提升57.68%的执行性能，同时降低32%~43%的互连开销.

Abstract

Reconfigurable arrays are widely used as accelerators in specific fields relying on the energy efficiency advantages brought by the driving of data flow. With the increase of application range

when there are regions with different execution rates in the application

it will cause serious performance loss by adopting the traditional spatial mapping scheme to map the whole data flow graph directly. A mapping method based on data flow decouplingisproposed to solve the mismatch of execution rate by adding decoupling units between regions with different execution rates. At the same time

a distributed multi-stage placement algorithm is used to map the decoupled dataflow graph into a clustered interconnection structure. Compared with the traditional mapping scheme

the mapping scheme after decoupling the data flow can improve the execution performance by 57.68% on average while maintaining a high interconnection routabilityand reduce interconnection costs by 32% to 43%.

关键词

Keywords

references

ZHAO Z Y, LIU Y T, SHENG W G, et al. Optimizing the data placement and transformation for multi-bank CGRA computing system[C]//2018 Design, Automation % Test in Europe Conference % Exhibition (DATE). Dresden, Germany: IEEE, 2018: 1087-1092. DOI: 10.23919/DATE.2018.8342173.

PRABHAKAR R, ZHANG Y Q, KOEPLINGER D, et al. Plasticine: A reconfigurable architecture for parallel patterns[C]//2017 ACM/IEEE 44th Annual International Symposium on Computer Architecture (ISCA). Toronto, ON, Canada: IEEE, 2017: 389-402. DOI: 10.1145/3079856.3080256.

CONG J, HUANG H, MA C Y, et al. A fully pipelined and dynamically composable architecture of CGRA[C]//2014 IEEE 22nd Annual International Symposium on Field-Programmable Custom Computing Machines. Boston, MA, USA: IEEE, 2014: 9-16. DOI: 10.1109/FCCM.2014.12.

HONG T, GUAN N, YIN C, et al. Decoupling the multi-rate dataflow execution in coarse-grained reconfigurable array[C]//2020 IEEE International Symposium on Circuits and Systems (ISCAS). Seville, Spain: IEEE, 2020: 1-5. DOI: 10.1109/ISCAS45731.2020.9181275.

New York BETZ V, ROSE J, MARQUARDT A. Architecture and CAD for deep-submicron FPGAs[M]. New York: Springer, 1999: 51-55. DOI: 51

DE P, MANDAL C, PRAMPALLI U. Path-balanced logic design to realize block ciphers resistant to power and timing attacks[J]. IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2019, 27(5): 1080-1092. DOI: 10.1109/TVLSI.2019.2896377.

REAGEN B, ADOLF R, SHAO Y S, et al. MachSuite: Benchmarks for accelerator design and customized architectures[C]//2014 IEEE International Symposium on Workload Characterization (ⅡSWC). Raleigh, NC, USA: IEEE, 2014: 110-119. DOI: 10.1109/ⅡSWC.2014.6983050.

ROSENFELD P, COOPER-BALIS E, JACOB B. DRAMSim2: A cycle accurate memory system simulator[J]. IEEE Computer Architecture Letters, 2011, 10(1): 16-19. DOI: 10.1109/L-CA.2011.4.

MEI B F, VERNALDE S, VERKEST D, et al. ADRES: An architecture with tightly coupled VLIW processor and coarse-grained reconfigurable matrix[C]//International Conference on Field Programmable Logic and Applications. Berlin, Heidelberg: Springer, 2003: 61-70. DOI: 10.1007/978-3-540-45234-8_7.

MIRHOSEINI A, GOLDIE A, YAZGAN M, et al. Chip placement with deep reinforcement learning[J]. arXiv e-prints, 2020: arXiv: 2004.10746.

LIN X H, YIN S Y, LIU L B, et al. Exploiting parallelism of imperfect nested loops with sibling inner loops on coarse-grained reconfigurable architectures[C]//2016 21st Asia and South Pacific Design Automation Conference (ASP-DAC). Macao, China: IEEE, 2016: 456-461. DOI: 10.1109/ASPDAC.2016.7428054.

VOITSECHOV D, ETSION Y. Single-graph multiple flows: Energy efficient design alternative for GPGPUs[J]. ACM SIGARCH Computer Architecture News, 2014, 42(3): 205-216. DOI: 10.1145/2678373.2665703.

YIN S Y, YAO X Q, LU T Y, et al. Conflict-free loop mapping for coarse-grained reconfigurable architecture with multi-bank memory[J]. IEEE Transactions on Parallel and Distributed Systems, 2017, 28(9): 2471-2485. DOI: 10.1109/TPDS.2017.2682241.

SINGH H, LEE M H, LU G M, et al. MorphoSys: an integrated reconfigurable system for data-parallel and computation-intensive applications[J]. IEEE Transactions on Computers, 2000, 49(5): 465-481. DOI: 10.1109/12.859540.

行华彧, 景乃锋. 一种基于多阶段模拟退火的异构可重构阵列布局算法[J]. 微电子学与计算机, 2020, 37(6): 1-5. DOI: 1 XING H Y, JING N F. A placement algorithm for HGRA based on multi-stage simulatedanneal[J]. Microelectronics %Computer, 2020, 37(6): 1-5. DOI: 1

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

低功耗设计的双链电源关断策略的实现与验证

大模型发展综述：从BERT到Deepseek的技术演进

基于反向体偏置的高增益稳定性高温运放设计

基于区块链的AI生成内容溯源框架设计研究

超低功耗实时时钟电路的设计