Uses a set of registers to implement a leaky integrate-and-fire (LIF) neuron for spike-timing dependent plasticity learning (STDP) rule. Two LIF neurons are instantiated and a stdp module handles the logic for the timing and weight update.
After reset, a current is applied at different amplitudes and the input to the neuron is integrated at every clock cycle If a pre-synaptic spike and post-synaptic spike occurs, time difference is measured and applied to the synatic weight update for this connection.
| # | Input | Output | Bidirectional |
|---|---|---|---|
| 0 | drive current | LIF spike | second counter bit 0 |
| 1 | LIF state | second counter bit 1 | |
| 2 | synaptic weight | second counter bit 2 | |
| 3 | second counter bit 3 | ||
| 4 | second counter bit 4 | ||
| 5 | second counter bit 5 | ||
| 6 | second counter bit 6 | ||
| 7 | second counter bit 7 |