In 1971 Leon Chua, an American electrical engineer, described a hypothetical device called a memristor. The combination of the words memory and resistor gives a clue to its purpose; that being the ability to remember its electrical history. This resistor would remember all the current that had flowed through it in the past and this would affect its current resistivity properties. It has been estimated that the efficient production of memristors could be complete by 2019 but to meet this target breakthroughs need to happen to push the possible but difficult and expensive designs into a more practical format.
And so experiments in the quantum arena have been performed and designs drawn up. The idea is to create a feedback circuit involving superconductors (although it could be theoretically done without them) to retain the memory all submerged in a Markovian bath. This word Markovian is used when a system is unaffected by its past states. When all this is placed into a component it is a quantum memristor which is of course non-Markovian. The development of these devices heralds more than just a classical advance in computer memory of power. It means a computer can evolve. These components change based past events and it is hard not to make a comparison to neurons in the brain and their ability to reorganise and adapt based on situation. Since the ultimate goal of a computer is to match a human brain this is possibly the next technological step to doing so.