Real-time scalable cortical computing at 46 giga-synaptic OPS/watt with ~100× speedup in time-to-solution and ~100,000× reduction in energy-to-solution

摘要: Drawing on neuroscience, we have developed a parallel, event-driven kernel for neurosynaptic computation, that is efficient with respect to memory, and communication. Building the previously demonstrated highly optimized software expression of kernel, here, demonstrate True North, co-designed silicon kernel. North achieves five orders magnitude reduction in energy to-solution two speedup time-to solution, when running computer vision applications complex recurrent neural network simulations. Breaking path von Neumann architecture, 4,096 core, 1 million neuron, 256 synapse brain-inspired processor, consumes 65mW power at real-time delivers performance 46 Giga-Synaptic OPS/Watt. We seamless tiling chips into arrays, forming foundation cortex-like scalability. North's unprecedented time-to-solution, energy-to-solution, size, scalability, combined underlying flexibility enable broad range cognitive applications.