Furthermore, an artificial network is constructed based on our device parameters and revealed a stunning learning accuracy of 94% for the MNIST handwritten datasets.
The device parameters are compared with the reported values to establish its unique adaptive behavior at such a low stimulus, like biological synapses. Excellent switching features with spike-time, spike-rate-dependent plasticity, long-term potentiation, and depression are demonstrated within 500 mV. Herein, it is shown that Na modified nanocrystalline molybdenum disulfide (MoS2) flexible memristors can provide great feasibility in ultra-low-voltage switching operations as well as synaptic behavior. This provides the inspiration for the current memristive technology to reach the ubiquitous goal of concocting neuromorphic electronics.
The human brain comprises 10 15 synapses and consumes only 20W of power, where a single synaptic function requires an activation potential of around 100 mV.
0 kommentar(er)
