A 65k-neuron 73-Mevents/s 22-pJ/event asynchronous micro-pipelined integrate-and-fire array transceiver

Jongkil Park, Sohmyung Ha, Theodore Yu, Emre Neftci, Gert Cauwenberghs

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    We present a 65k-neuron integrate-and-fire array transceiver (IFAT) for spike-based neural computation with low-power, high-throughput connectivity. The internally analog, externally digital chip is fabricated on a 4×4 mm2 die in 90nm CMOS and arranged in 4 quadrants of 16k parallel addressable neurons. Each neuron circuit serves input spike events by dynamically instantiating conductance-based synapses onto four local synapse circuits over two membrane compartments, and produces output spike events upon reaching a threshold in integration over one of the membrane compartments. Fully asynchronous input and output spike event data streams are mediated over the standard address event representation (AER) protocol. To support full event throughput at large synaptic fan-in, a two-tier micro-pipelining scheme parallelizes input events along neural array cores, and along rows of each core. Measured results show sustained peak synaptic event throughput of 18.2 Mevents/s per quadrant, at 22 pJ average energy per synaptic input event and 25 μW standby power.

    Original languageEnglish (US)
    Title of host publicationIEEE 2014 Biomedical Circuits and Systems Conference, BioCAS 2014 - Proceedings
    PublisherInstitute of Electrical and Electronics Engineers Inc.
    Pages675-678
    Number of pages4
    ISBN (Electronic)9781479923465
    DOIs
    StatePublished - Jan 1 2014
    Event10th IEEE Biomedical Circuits and Systems Conference, BioCAS 2014 - Lausanne, Switzerland
    Duration: Oct 22 2014Oct 24 2014

    Other

    Other10th IEEE Biomedical Circuits and Systems Conference, BioCAS 2014
    CountrySwitzerland
    CityLausanne
    Period10/22/1410/24/14

    Fingerprint

    Transceivers
    Neurons
    Fires
    Throughput
    Membranes
    Networks (circuits)

    ASJC Scopus subject areas

    • Hardware and Architecture
    • Biomedical Engineering

    Cite this

    Park, J., Ha, S., Yu, T., Neftci, E., & Cauwenberghs, G. (2014). A 65k-neuron 73-Mevents/s 22-pJ/event asynchronous micro-pipelined integrate-and-fire array transceiver. In IEEE 2014 Biomedical Circuits and Systems Conference, BioCAS 2014 - Proceedings (pp. 675-678). [6981816] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/BioCAS.2014.6981816

    A 65k-neuron 73-Mevents/s 22-pJ/event asynchronous micro-pipelined integrate-and-fire array transceiver. / Park, Jongkil; Ha, Sohmyung; Yu, Theodore; Neftci, Emre; Cauwenberghs, Gert.

    IEEE 2014 Biomedical Circuits and Systems Conference, BioCAS 2014 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2014. p. 675-678 6981816.

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Park, J, Ha, S, Yu, T, Neftci, E & Cauwenberghs, G 2014, A 65k-neuron 73-Mevents/s 22-pJ/event asynchronous micro-pipelined integrate-and-fire array transceiver. in IEEE 2014 Biomedical Circuits and Systems Conference, BioCAS 2014 - Proceedings., 6981816, Institute of Electrical and Electronics Engineers Inc., pp. 675-678, 10th IEEE Biomedical Circuits and Systems Conference, BioCAS 2014, Lausanne, Switzerland, 10/22/14. https://doi.org/10.1109/BioCAS.2014.6981816
    Park J, Ha S, Yu T, Neftci E, Cauwenberghs G. A 65k-neuron 73-Mevents/s 22-pJ/event asynchronous micro-pipelined integrate-and-fire array transceiver. In IEEE 2014 Biomedical Circuits and Systems Conference, BioCAS 2014 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2014. p. 675-678. 6981816 https://doi.org/10.1109/BioCAS.2014.6981816
    Park, Jongkil ; Ha, Sohmyung ; Yu, Theodore ; Neftci, Emre ; Cauwenberghs, Gert. / A 65k-neuron 73-Mevents/s 22-pJ/event asynchronous micro-pipelined integrate-and-fire array transceiver. IEEE 2014 Biomedical Circuits and Systems Conference, BioCAS 2014 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2014. pp. 675-678
    @inproceedings{f98de974301340cc978e311bb749856b,
    title = "A 65k-neuron 73-Mevents/s 22-pJ/event asynchronous micro-pipelined integrate-and-fire array transceiver",
    abstract = "We present a 65k-neuron integrate-and-fire array transceiver (IFAT) for spike-based neural computation with low-power, high-throughput connectivity. The internally analog, externally digital chip is fabricated on a 4×4 mm2 die in 90nm CMOS and arranged in 4 quadrants of 16k parallel addressable neurons. Each neuron circuit serves input spike events by dynamically instantiating conductance-based synapses onto four local synapse circuits over two membrane compartments, and produces output spike events upon reaching a threshold in integration over one of the membrane compartments. Fully asynchronous input and output spike event data streams are mediated over the standard address event representation (AER) protocol. To support full event throughput at large synaptic fan-in, a two-tier micro-pipelining scheme parallelizes input events along neural array cores, and along rows of each core. Measured results show sustained peak synaptic event throughput of 18.2 Mevents/s per quadrant, at 22 pJ average energy per synaptic input event and 25 μW standby power.",
    author = "Jongkil Park and Sohmyung Ha and Theodore Yu and Emre Neftci and Gert Cauwenberghs",
    year = "2014",
    month = "1",
    day = "1",
    doi = "10.1109/BioCAS.2014.6981816",
    language = "English (US)",
    pages = "675--678",
    booktitle = "IEEE 2014 Biomedical Circuits and Systems Conference, BioCAS 2014 - Proceedings",
    publisher = "Institute of Electrical and Electronics Engineers Inc.",

    }

    TY - GEN

    T1 - A 65k-neuron 73-Mevents/s 22-pJ/event asynchronous micro-pipelined integrate-and-fire array transceiver

    AU - Park, Jongkil

    AU - Ha, Sohmyung

    AU - Yu, Theodore

    AU - Neftci, Emre

    AU - Cauwenberghs, Gert

    PY - 2014/1/1

    Y1 - 2014/1/1

    N2 - We present a 65k-neuron integrate-and-fire array transceiver (IFAT) for spike-based neural computation with low-power, high-throughput connectivity. The internally analog, externally digital chip is fabricated on a 4×4 mm2 die in 90nm CMOS and arranged in 4 quadrants of 16k parallel addressable neurons. Each neuron circuit serves input spike events by dynamically instantiating conductance-based synapses onto four local synapse circuits over two membrane compartments, and produces output spike events upon reaching a threshold in integration over one of the membrane compartments. Fully asynchronous input and output spike event data streams are mediated over the standard address event representation (AER) protocol. To support full event throughput at large synaptic fan-in, a two-tier micro-pipelining scheme parallelizes input events along neural array cores, and along rows of each core. Measured results show sustained peak synaptic event throughput of 18.2 Mevents/s per quadrant, at 22 pJ average energy per synaptic input event and 25 μW standby power.

    AB - We present a 65k-neuron integrate-and-fire array transceiver (IFAT) for spike-based neural computation with low-power, high-throughput connectivity. The internally analog, externally digital chip is fabricated on a 4×4 mm2 die in 90nm CMOS and arranged in 4 quadrants of 16k parallel addressable neurons. Each neuron circuit serves input spike events by dynamically instantiating conductance-based synapses onto four local synapse circuits over two membrane compartments, and produces output spike events upon reaching a threshold in integration over one of the membrane compartments. Fully asynchronous input and output spike event data streams are mediated over the standard address event representation (AER) protocol. To support full event throughput at large synaptic fan-in, a two-tier micro-pipelining scheme parallelizes input events along neural array cores, and along rows of each core. Measured results show sustained peak synaptic event throughput of 18.2 Mevents/s per quadrant, at 22 pJ average energy per synaptic input event and 25 μW standby power.

    UR - http://www.scopus.com/inward/record.url?scp=84920548116&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84920548116&partnerID=8YFLogxK

    U2 - 10.1109/BioCAS.2014.6981816

    DO - 10.1109/BioCAS.2014.6981816

    M3 - Conference contribution

    SP - 675

    EP - 678

    BT - IEEE 2014 Biomedical Circuits and Systems Conference, BioCAS 2014 - Proceedings

    PB - Institute of Electrical and Electronics Engineers Inc.

    ER -