High Bandwidth Memory on FPGAs: A Data Analytics Perspective

被引:24
|
作者
Kara, Kaan [1 ,2 ]
Hagleitner, Christoph [2 ]
Diamantopoulos, Dionysios [2 ]
Syrivelis, Dimitris [2 ]
Alonso, Gustavo [1 ]
机构
[1] Swiss Fed Inst Technol, Dept Comp Sci, Syst Grp, Zurich, Switzerland
[2] IBM Res Europe, Zurich, Switzerland
关键词
High Bandwidth Memory (HBM); FPGA; Database; Advanced Analytics;
D O I
10.1109/FPL50879.2020.00013
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
FPGA-based data processing in datacenters is increasing in popularity due to the demands of modern workloads and the resulting need for specialization in hardware. Driven by this trend, vendors are rapidly adapting reconfigurable devices to suit data and compute intensive workloads. Inclusion of High Bandwidth Memory (HBM) in FPGA devices is a recent example. HBM promises overcoming the bandwidth bottleneck, often faced by FPGA-based accelerators due to their throughput oriented design. In this paper, we study the usage and benefits of HBM on FPGAs from a data analytics perspective. We consider three workloads that are often performed in analytics oriented databases and implement them on FPGA showing in which cases they benefit from HBM: range selection, hash join, and stochastic gradient descent for linear model training. We integrate our designs into a columnar database (MonetDB) and show the trade-offs arising from the integration related to data movement and partitioning. In certain cases, FPGA+HBM based solutions are able to surpass the highest performance provided by either a 2-socket POWER9 system or a 14-core XeonE5 by up to 1.8x (selection), 12.9x (join), and 3.2x (SGD).
引用
收藏
页码:1 / 8
页数:8
相关论文
共 50 条
  • [1] Shuhai: Benchmarking High Bandwidth Memory on FPGAs
    Wang, Zeke
    Huang, Hongjing
    Zhang, Jie
    Alonso, Gustavo
    [J]. 28TH IEEE INTERNATIONAL SYMPOSIUM ON FIELD-PROGRAMMABLE CUSTOM COMPUTING MACHINES (FCCM), 2020, : 111 - 119
  • [2] Shuhai: A Tool for Benchmarking High Bandwidth Memory on FPGAs
    Huang, Hongjing
    Wang, Zeke
    Zhang, Jie
    He, Zhenhao
    Wu, Chao
    Xiao, Jun
    Alonso, Gustavo
    [J]. IEEE TRANSACTIONS ON COMPUTERS, 2022, 71 (05) : 1133 - 1144
  • [3] Quantifying effective memory bandwidth of platform FPGAs
    Schmidt, Andrew G.
    Sass, Ron
    [J]. FCCM 2007: 15TH ANNUAL IEEE SYMPOSIUM ON FIELD-PROGRAMMABLE CUSTOM COMPUTING MACHINES, PROCEEDINGS, 2007, : 337 - +
  • [4] Accelerating Next Generation Genome Sequencing Leveraging High Bandwidth Memory on FPGAs
    Lehmann, Till
    Wenzel, Lukas
    Plauth, Max
    Koehler, Sven
    Polze, Andreas
    [J]. 2022 TENTH INTERNATIONAL SYMPOSIUM ON COMPUTING AND NETWORKING WORKSHOPS, CANDARW, 2022, : 62 - 68
  • [5] StreamBox-HBM: Stream Analytics on High Bandwidth Hybrid Memory
    Miao, Hongyu
    Jeon, Myeongjae
    Pekhimenko, Gennady
    McKinley, Kathryn S.
    Lin, Felix Xiaozhu
    [J]. TWENTY-FOURTH INTERNATIONAL CONFERENCE ON ARCHITECTURAL SUPPORT FOR PROGRAMMING LANGUAGES AND OPERATING SYSTEMS (ASPLOS XXIV), 2019, : 167 - 181
  • [6] In-Network Online Data Analytics with FPGAs
    Cooke, Ryan
    Fahmy, Suhaib A.
    [J]. 2017 27TH INTERNATIONAL CONFERENCE ON FIELD PROGRAMMABLE LOGIC AND APPLICATIONS (FPL), 2017,
  • [7] Accelerating Big Data Analytics Using FPGAs
    Neshatpour, Katayoun
    Malik, Maria
    Ghodrat, Mohammad Ali
    Homayoun, Houman
    [J]. 2015 IEEE 23RD ANNUAL INTERNATIONAL SYMPOSIUM ON FIELD-PROGRAMMABLE CUSTOM COMPUTING MACHINES (FCCM), 2015, : 164 - 164
  • [8] Enhancing Memory Bandwidth in a Single Stream Computation with Multiple FPGAs
    Mondigo, Antoniette
    Sano, Kentaro
    Takizawa, Hiroyuki
    [J]. 2018 INTERNATIONAL CONFERENCE ON FIELD-PROGRAMMABLE TECHNOLOGY (FPT 2018), 2018, : 381 - 383
  • [9] MEG: A RISCV-based System Emulation Infrastructure for Near-data Processing Using FPGAs and High-bandwidth Memory
    Zhang, Jialiang
    Zha, Yue
    Beckwith, Nicholas
    Liu, Bangya
    Li, Jing
    [J]. ACM TRANSACTIONS ON RECONFIGURABLE TECHNOLOGY AND SYSTEMS, 2020, 13 (04)
  • [10] A Scalable Emulator for Quantum Fourier Transform Using Multiple-FPGAs with High-Bandwidth-Memory
    Waidyasooriya, Hasitha Muthumala
    Oshiyama, Hiroki
    Kurebayashi, Yuya
    Hariyama, Masanori
    Ohzeki, Masayuki
    [J]. IEEE ACCESS, 2022, 10 : 65103 - 65117