Skip to main content
Video s3
    Details
    Presenter(s)
    Gopabandhu Hota Headshot
    Display Name
    Gopabandhu Hota
    Affiliation
    Affiliation
    University of California, San Diego
    Country
    Author(s)
    Display Name
    Gopabandhu Hota
    Affiliation
    Affiliation
    University of California, San Diego
    Display Name
    Nishant Mysore
    Affiliation
    Affiliation
    University of California, San Diego
    Display Name
    Stephen Deiss
    Affiliation
    Affiliation
    University of California, San Diego
    Display Name
    Bruno Pedroni
    Affiliation
    Affiliation
    University of California San Diego
    Display Name
    Gert Cauwenberghs
    Affiliation
    Affiliation
    University of California, San Diego
    Abstract

    We present a high-performance and low-overhead multicast network-on-chip (NoC) architecture for hierarchical address event routing (Multicast-HiAER) suitable for large-scale reconfigurable neuromorphic systems. Each building block of this efficient NoC architecture consists of several multi-cast advanced high-performance buses (mAHB) running in parallel for high-bandwidth inter-core spike event transmission. This architecture for scalable event routing can help to implement brain-scale neural network connectivity distributed across neuromorphic processing cores, with network constraints typical of locally dense and globally sparse neuron connectivity. For a demonstration using a Xilinx Virtex Ultrascale VU37p FPGA, we have shown an 8$times$8 grid of mAHBs running at 512MHz clock performing Level-1 and Level-2 inter-core communication at top bandwidth of 420M events per second per 128k neuron node in the hierarchy. This peak absolute bandwidth supports spike event registration with sub-ms latencies under worst-case conditions of all postsynaptic destinations being off-core.

    Slides
    • Hierarchical Multicast Network-on-Chip for Scalable Reconfigurable Neuromorphic Systems (application/pdf)