Progress of knowledge eases the best way to entry the world however requires growing quantities of vitality to retailer and course of. Neuromorphic electronics has emerged within the final decade, impressed by organic neurons and synapses, with in-memory computing skill, extenuating the ‘von Neumann bottleneck’ between the reminiscence and processor and providing a promising answer to cut back the efforts each in knowledge storage and processing, due to their multi-bit non-volatility, biology-emulated traits, and silicon compatibility. This work opinions the current advances in rising memristive units for synthetic neuron and synapse functions, together with reminiscence and data-processing skill: the physics and traits are mentioned first, i.e., valence altering, electrochemical metallization, part altering, interfaced-controlling, charge-trapping, ferroelectric tunnelling, and spin-transfer torquing. Subsequent, we suggest a common benchmark for the bogus synapse and neuron units on spiking vitality consumption, standby energy consumption, and spike timing. Primarily based on the benchmark, we deal with the challenges, recommend the rules for intra-device and inter-device design, and supply an outlook for the neuromorphic functions of resistive switching-based synthetic neuron and synapse units.