Neuromorphic chip (Intel Loihi / IBM TrueNorth)

Realizes: spiking neural network computation

Silicon chips that mimic neural computation using spiking neurons and synaptic connections. Intel Loihi and IBM TrueNorth implement event-driven, asynchronous processing with on-chip learning capabilities. Speed: microseconds (spike propagation). Capacity: millions of neurons (parallel event-driven processing).

Examples

🔗

Intel Loihi 2 Neuromorphic Chip

Intel's second-generation neuromorphic processor with 128 neuromorphic cores supporting 130,000 synthetic neurons and 130M synapses, 10x faster than Loihi 1

INTEGRATE-AND-FIRE WEIGHTED-SUM THRESHOLD seconds large pJ

🔗

IBM TrueNorth Neuromorphic Chip

IBM's neuromorphic processor with 4,096 neurosynaptic cores, 1 million programmable neurons and 256 million synapses, consuming only 70 milliwatts

INTEGRATE-AND-FIRE WEIGHTED-SUM THRESHOLD seconds large pJ

🔗

IBM NorthPole Neuromorphic Chip

IBM's 2023 neuromorphic processor running ResNet-50 22x faster and 25x more energy efficient than GPUs, 4000x faster than TrueNorth

INTEGRATE-AND-FIRE WEIGHTED-SUM THRESHOLD seconds large pJ