SpiNNaker

Realizes: Massively parallel ARM968 neuromorphic fabric for real-time spiking networks

SpiNNaker machines at the University of Manchester network over a million ARM968 cores via packet-switched triple-torus routers to run spiking neural networks with local plasticity and sensor-motor I/O in real time, letting cortical-scale models stay synchronized as spikes hop across the mesh.

Examples

🔗

Human Brain Project cortical microcircuit on SpiNNaker

A 1-million-core SpiNNaker board hosts Human Brain Project cortical microcircuit models, routing roughly a billion synapses in real time with tens of microsecond latencies.

Mapping PyNN cortical microcircuit descriptions onto SpiNNaker chips, routing spikes through the packet-switched mesh, applying STDP synaptic updates, and streaming sensor/motor I O in real time. real-time cluster-scale tens of picojoules per synaptic event

🔗

SpiNNaker 1 board

The SpiNNaker 1 board integrates 48 ARM968 cores, SDRAM, and the triple-torus routers used in Manchester machines, letting researchers map PyNN cortical networks and stream spikes through the on-board fabric for logging.

ROUTE_SPIKES SIMULATE_SPIKES STREAM_SPIKE_DATA real-time board-scale tens of picojoules per synaptic event