# Memristive Neuromorphic Computing
**Domain:** Neuromorphic Devices / In-Memory Computing
**Doc Type:** Technical Concept Node
**Maturity:** Active research and early hardware development
**Primary Source:** [[articles/The Organic-Synthetic Brain Atlas|The Organic-Synthetic Brain Atlas]]
## Definition
**Memristive neuromorphic computing** uses devices whose conductance depends on prior electrical history to represent and update synaptic-like weights in physical hardware. A memristor can combine storage and computation in the same element, reducing the data movement that dominates energy use in conventional processor-memory architectures.
## Architecture
In a crossbar array, conductances encode a matrix and applied voltages produce output currents through the array. The physical circuit performs a form of matrix-vector multiplication directly through Ohm's and Kirchhoff's laws. Diffusive and threshold-switching devices can also implement neuron-like integration and firing dynamics.
## Relation to the atlas
The organic-synthetic atlas places memristive devices beyond today's programmable neuromorphic chips on a device-level trajectory toward denser, lower-energy, and more locally adaptive hardware. They complement [[wiki/Spiking Neural Network|spiking neural networks]] but do not require every workload to be spiking.
## Constraints
Device variability, endurance, write noise, retention, fabrication yield, analog precision, and integration with conventional control electronics remain central limitations. A synapse-like electrical response is an engineering analogy, not evidence that the device reproduces the biological or experiential properties of a synapse.
## Relationships
**Related cluster nodes:** [[wiki/Neuromorphic Computing|Neuromorphic Computing]] · [[wiki/Spiking Neural Network|Spiking Neural Network]] · [[wiki/Photonic Neuromorphic Computing|Photonic Neuromorphic Computing]] · [[collections/Neurotech|Neurotech]]