RP-Ring: A Heterogeneous Multi-FPGA Accelerator

被引:2
|
作者
Guo, Shuaizhi [1 ]
Wang, Tianqi [1 ]
Tao, Linfeng [1 ]
Tian, Teng [1 ]
Xiang, Zikun [1 ]
Jin, Xi [1 ]
机构
[1] Univ Sci & Technol China, Hefei, Anhui, Peoples R China
关键词
D O I
10.1155/2018/6784319
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
To reduce the cost of designing new specialized FPGA boards as direct-summation MOND (Modified Newtonian Dynamics) simulator, we propose a new heterogeneous architecture with existing FPGA boards, which is called RP-ring (reconfigurable processor ring). This design can be expanded conveniently with any available FPGA board and only requires quite low communication bandwidth between FPGA boards. The communication protocol is simple and can be implemented with limited hardware/software resources. In order to avoid overall performance loss caused by the slowest board, we build a mathematical model to decompose workload among FPGAs. The dividing of workload is based on the logic resource, memory access bandwidth, and communication bandwidth of each FPGA chip. Our accelerator can achieve two orders of magnitude speedup compared with CPU implementation.
引用
收藏
页数:14
相关论文
共 50 条
  • [41] A parallel compact genetic algorithm for Multi-FPGA Partitioning
    Hidalgo, JI
    Baraglia, R
    Perego, R
    Lanchares, J
    Tirado, F
    NINTH EUROMICRO WORKSHOP ON PARALLEL AND DISTRIBUTED PROCESSING, PROCEEDINGS, 2001, : 113 - 120
  • [42] Demonstration of Flow-in-Cloud: A multi-FPGA system
    Hironaka, Kazuei
    Iizuka, Kensuke
    Ben Ahmed, Akram
    Ullah, M. M. Imdad
    Yamauchi, Yugo
    Sun, Yuxi
    Yamakura, Miho
    Hiruma, Aoi
    Amano, Hideharu
    2019 29TH INTERNATIONAL CONFERENCE ON FIELD-PROGRAMMABLE LOGIC AND APPLICATIONS (FPL), 2019, : 417 - 418
  • [43] HitGNN: High-Throughput GNN Training Framework on CPU plus Multi-FPGA Heterogeneous Platform
    Lin, Yi-Chien
    Zhang, Bingyi
    Prasanna, Viktor K.
    IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, 2024, 35 (05) : 707 - 719
  • [44] Multi-FPGA Partitioning Method Based on Topological Levelization
    Kerkiz, Nabil
    Elchouemi, Amr
    Bouldin, Don
    JOURNAL OF ELECTRICAL AND COMPUTER ENGINEERING, 2010, 2010
  • [45] Two IP Protection Schemes for Multi-FPGA Systems
    Gaspar, Lubos
    Fischer, Viktor
    Gueneysu, Tim
    Cherif Jouini, Zouha
    2012 INTERNATIONAL CONFERENCE ON RECONFIGURABLE COMPUTING AND FPGAS (RECONFIG), 2012,
  • [46] Integrating HDL synthesis and partitioning for multi-FPGA designs
    Fang, WJ
    Wu, ACH
    IEEE DESIGN & TEST OF COMPUTERS, 1998, 15 (02): : 65 - 72
  • [47] A Design Workflow for Dynamically Reconfigurable Multi-FPGA Systems
    Panella, Alessandro
    Santambrogio, Marco D.
    Redaelli, Francesco
    Cancare, Fabio
    Sciuto, Donatella
    PROCEEDINGS OF THE 2010 18TH IEEE/IFIP INTERNATIONAL CONFERENCE ON VLSI AND SYSTEM-ON-CHIP, 2010, : 414 - 419
  • [48] INTEGRATION OF A MULTI-FPGA SYSTEM IN A COMMON CLUSTER ENVIRONMENT
    Knodel, Oliver
    Spallek, Rainer G.
    2013 23RD INTERNATIONAL CONFERENCE ON FIELD PROGRAMMABLE LOGIC AND APPLICATIONS (FPL 2013) PROCEEDINGS, 2013,
  • [49] An efficient partitioning method with UCF for a multi-FPGA system
    Zhang, Q.-L. (zhangqianli@ime.ac.cn), 1600, Harbin Institute of Technology (44):
  • [50] Lattice-based Scheduling for Multi-FPGA Systems
    Yu, Teng
    Feng, Bo
    Stillwell, Mark
    Guo, Liucheng
    Ma, Yuchun
    Thomson, John
    2018 INTERNATIONAL CONFERENCE ON FIELD-PROGRAMMABLE TECHNOLOGY (FPT 2018), 2018, : 321 - 324