In a paper published in Wednesday's edition of Nature magazine, four researchers from HP 's Information and Quantum System Labs discussed their creation of a mathematical model and physical example of what is called a "memristor," a conjoined word formed from "memory" and "resistor."

Similar to Human Brain

The researchers said memristors can lead to computer systems with memories that are not erased when power is lost, do not need to be booted up, are very low-power, and, in another encroachment on the uniqueness of humans, associate information "in a manner similar to that of the human brain."

The possibility of such an electronic component was first theorized nearly 40 years ago by Leon Chua, a professor in the electrical engineering and computer sciences department at the University of California, Berkeley.

Chua said there was a fourth fundamental circuit element, the memristor, to accompany the existing three -- resistor, capacitor and inductor. He also theorized that its properties could not be obtained by any combination of the other three. Like chemical elements, fundamental circuit elements cannot be created from the others.

R. Stanley Williams, head of the HP team, said the research will make it "possible for engineers to develop integrated circuit designs that could dramatically improve the performance and energy efficiency " for PCs and data centers.

At Least Five Years

The researchers also suggested that memristor-based computers could help the momentum toward "cloud computing," in which applications and data are stored throughout many servers in-house, or on the Internet. Memristors could mean much less power to drive that infrastructure , and a greater resiliency in a data center if power is interrupted.

Memristors could lead to a new kind of computer memory that could complement or even replace the common DRAM (dynamic random-access memory), which does not store information when power is turned off.

Memristor-based memory can also lead to fundamentally different ways of using computer memory. The researchers said memristor technology can recall and make associations between events that are similar to how a human brain recognizes patterns.

In fact, Chua's original paper proposed that memristors would act in ways that are similar to how synapses work in a brain. One of the ways is that the more often a signal is sent to a synapse, which connects two neurons, the stronger the synapse becomes.

The HP team has told news media that this enhanced ability to deal with patterns could allow computers to makes decisions based on past data. This might include, for instance, the development of a more intelligent microwave that knows the heating times that you prefer for different kinds of foods. But it won't happen overnight. The first commercial uses are not expected for at least five years.