Although the computing power of computers is continually increasing, they are and continue to be clumsy energy wasters when compared to human brains. however, researchers have developed a building block that functions in the same manner as a synapse in the brain and is just as energy efficient as the biological example.

The basic element of modern micro-electronics is the transistor; but this has a few disadvantages. It only knows 2 states, so that many are required for the storage or processing of data. It also requires a continuous supply of current to prevent memory loss. Finally, storage and processing operations are separated from each other, so that data often has to be transported back and forth.

In our brains this works very differently. The smallest element there is the synapse, and this knows a large number of 'states', does not require a continuous energy supply and is close to the data-processing neurons. As a consequence our brains can efficiently process a large amount of information at the same time (for example consider the recognition of faces and objects).

The new building block that has been developed by a group of researchers at Stanford University, has none of the disadvantages of the classic transistor. The artificial synapse (which could form the basis of an artificial brain) comprises three layers and has some resemblance to a battery - the bottom and top layers are the 'poles', with an electrolyte in between. The composition, and with that the resistance, of the bottom pole can be changed by applying current to the top pole and can have more than 500 different values. This mechanism can be used for non-volatile data storage (the information remains when the current is removed). The artificial synapse is made from inexpensive and flexible polymers, organic materials that should make it possible some day to made direct data connections with the brain.

A summary of the research paper can be found here.