Memristors as electronic synapses


Perhaps the most important news in science of the past week was the rediscovery of memristors (Nature, May 2008), a basic passive electronic element that supplements the conventional three passive circuit elements: the resistor, the capacitor, and the inductor. A memristor is a resistor whose resistance changes as the function of the current that has flown through the element.

The main memristor application everyone seems to focus on is memory: cheaper than existing transistor-based memory, less power-hungry, and nonvolatile. Memristor-based devices will not need to be booted at startup and will not need to consume energy to maintain memory contents.

However, memristors will also enable electronic circuits that combine memory and logic similarly to the way biological neural networks do. In classical computers, even in today’s artificial neural networks, memory and logic are physically separate. Erasing this artificial divide could lead not only to enormous increases in performance but to the emergence of new computational paradigms: malleable, adaptive, and dynamic.

Creating these technologies will require new computational abstractions. Much of today’s electrical engineering training is based on Linear Time-invariant (LTI) System Theory to characterize circuits. The addition of memristors into the mix prevents the use of current powerful abstractions such as the Laplace Transform, for example. Perhaps insight could be gained from studying biological systems. The nervous system does not separate function from memory and learning is accomplished by gradual modulation of connections’ strengths.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: