Home▸AI & Machine Learning▸Neuromorphic vs Clocked Processing 2D: Spike Raster & Energy Trace

Neuromorphic vs Clocked Processing 2D: Spike Raster & Energy Trace

2D companion to the neuromorphic-vs-clocked energy simulator: a scrolling spike raster and an analog membrane-potential trace, computed independently with the same leaky-integrate-and-fire and fixed-clock-sampling math, showing exactly why event-driven hardware saves energy on sparse sensor data.

AI & Machine Learning2DModerate60 FPS📱 Mobile-adapted⇄ 3D version

This is the 2D companion to the neuromorphic-vs-clocked-processing energy simulator, computed independently rather than reprojected from the 3D scene. Two 8×8 arrays of processing nodes still receive the identical sparse sensor-event stream, but instead of a rotating spatial grid of spheres, this page renders the phenomenon the way spike-train data is actually visualized in neuroscience and neuromorphic-hardware papers: a scrolling raster plot showing every discrete firing event per node over time, and an analog membrane-potential trace for a single representative node so you can watch the leaky-integrate-and-fire accumulator climb and reset in real time. A live predicted-vs-measured spike-rate readout demonstrates numerically that the simulation's long-run behaviour matches the closed-form rate formula, while the energy meters carry over the same activity-proportional-vs-fixed-cost accounting that explains why event-driven neuromorphic hardware saves power at the edge.

⚙ Under the hood

A scrolling spike raster and an analog membrane-potential trace, computed independently with the same leaky-integrate-and-fire and fixed-clock-sampling math as the 3D version, plus a live predicted-vs-measured spike-rate readout that shows the energy-per-spike advantage of event-driven neuromorphic hardware numerically.

neuromorphicedge computingspiking neural networkenergy efficiencyevent-drivenspike raster

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install