Researchers at University of Michigan have found that memristor can "learn" using same methods that human brain. To recall, memristor join resistance and a memory cell, and are considered fourth passive electrical parts. It is not yet using customer product, but new research is promising.

Scientists have displayed that there were similarities among memristor and synapses, making these compounds interesting for systems attempting to emulate brain. Indeed, in humans, learning occurs when simultaneous voltage spikes are generated by sensors. E.g. eyes see something or skin touches an object. Synapses then react by adapting to information get.

The results will be published in the journal Nano Letter’s show that the memristor behave identically to synapses that receives spikes of tension. Using network of memristor could thus give rise, theoretically, to supercomputers able of emulating certain brain functions according to scientists who were part funded by Agency for advanced research projects in defense (DARPA) of U.S. government.