It started with an announcement from HP Labs in May 2008. Scientists there had experimental proof that memristors, hypothesized by Dr. Leon Chua in 1971, existed.^[1]  By June 2008, the HP scientists were announcing that they had successfully engineered control over test devices.^[2]

In its way, it's a quintessential science story as it developed on two parallel tracks. There's something called the 'holy trinity' in electrical engineering. It has to do with integrated circuits and voltage characteristics. The 'holy trinity' are: inductors, resistors and capacitors.^[3]

In 1971, Dr. Chua, professor at the University of California's (Berkeley) Electrical Engineering and Computer Sciences Department, observed "an unfilled symmetry between fundamental electromagnetic equations relating charge and flux and their corresponding passive circuit elements."^[4]

In other words, from a mathematical perspective something seemed to be missing. Chua theorized that there was a fourth member of the 'holy trinity', a memory resistor or memristor. No one made the connection between Chua's hypothesis and the anomalies noted in the literature since the 1950s. 

It was R. Stanley Williams' work in molecular electronics at the HP Labs that led his colleague Greg Snider to look back in the literature seeking some theoretical explanation for bizarre voltage occurrences they were encountering. That's when Snider found the memristor hypothesis and brought it to Williams. "It was several years of scratching my head and thinking about it," before Williams realized that their molecular devices were really memristors. "It just hit me between the eyes."^[5] Williams, Snider and colleagues, Dmitri Strukov and Duncan Stewart went on to prove the existence of Dr. Chua's memristor, a circuit element that, unlike any other member of the 'holy trinity', remembers voltage characteristics.

Just months after the experimental proof announcement, another HP Labs team, this time led by Duncan Stewart announced that they'd achieved engineering control of a memristor. They built a switch--a memristor switch--that's 50 nanometres by 50 nanometres (50 billionths of a metre x 50 billionths of a metre). It contained a layer of a chemical most people know as a component of sunscreen or white paint, titanium dioxide. One of the team members, Jianhua Yang then discovered that manipulating the oxygen atoms in the layer would allow him an unprecedented level of control over the device.^[6] (To get an idea about how difficult that would be, under Jump points, click Sticky and fast.)

A memristor switch can be operated in either digital or analogue modes. In its digital state, the memristor's ability to retain information about how much and when current had been flowing has some interesting applications in the short term. A memristor device in its digital state could replace the current solid state memory (Flash) found in computers for example, with less expensive nonvolatile random access memory (RAM). For the user, it means no more rebooting when starting up the computer, every program would be instantly available and much less energy would be consumed.^[7]

Given that a converntional switch is digital, either 'on' or 'off' or '1' or '0', being able to dial, in its analogue mode, to values such as 0.3 or 1.8 means the memristor switch has a unique 'in-between' state that could allow it to process information the way a brain does. In the long term and in its analogue state, the memristor could lead to computers that learn. Typically, computers that learn do so via their software. With the memristor, it's the hardware that would do the learning but it would be a specific kind of learning. The memristor would allow pattern-matching which is only one of a human's brain's learning functions. It's possible that the pattern-matching could be used for facial and voice recognition applications.^[8]

Jump back

Sticky and fast

Jump points

See me, feel me