It was 1965 when Gordon Moore (co-founder of Intel) made the observation that the number of transistors per square inch of integrated circuits roughly doubled each year, starting from the time of introduction of these devices. What was in 1965 an observation become a law postulating that this trend would not come to a stop. Unfortunately, the limits of physics have caused a recent slowdown in the rhythm at which semiconductor companies can miniaturize transistors. Yesterday, an article on the NYT focused on a fresh announcement by Rice University and HP on a fundamental discovery able to overcome the barrier to the continued miniaturization of electronic components. Read the rest of this entry »

Neuromorphic engineers are studying the nervous system and trying to emulate its function and organization in their computational and robotics systems. They are hoping to match (or perhaps even exceed) the human brain in vision, hearing, pattern recognition and learning tasks (Boahen 2005). Read the rest of this entry »

Note: This posting summarizes some arguments I presented at the 2010 IEEE Symposium on Circuits and Systems. The complete presentation is available at this link.

In an earlier posting I presented arguments of why the idea of the memristor as a fourth fundamental circuit element is likely to be wrong. However, regardless of whether or not the memristor is considered as a fundamental circuit element, one may ask if it is technically correct to say that the researchers from HPLabs actually did discover a memristor. Read the rest of this entry »

After writing several memristor models with SPICE last week, I decided to post a short tutorial here on the subject. I’ll concentrate on modeling the memristor introduced in the Nature Nanoletters’ article Memristive Switching mechanism for metal/oxide/metal nanodevices by Yang et al., published in July 2008. More specifically the model is an enhanced version of the one derived in the article CNN Using Memristors for Neighborhood Connections (IEEE CNNA 2010) by me and Mika Laiho. Read the rest of this entry »

Insightful people have noted that the brain has been difficult to understand from simple measurements because of its relative irreducibility to single neuron dynamics, suspected to be a result of emergent properties.  Line up a few hundred million transistors very carefully and you get a fast but ’stupid’ processor; bag a bunch of proteins with a lipid membrane and you’re still far from a functioning cell; the list goes on.  Something about the whole, goes the cliche, is greater than the sum of its parts.  I’ve traditionally only heard about this concept as it applies to neuroscience, but I recently came across a great quote from D’Arcy Thompson that could use some fresh air on the Internet.  In terms of cell biology, he warns against thinking of a multicellular organism in terms of its constituent cells, arguing instead that the biophysics is better understood in terms of the whole tissue and the interactions between its pieces.  I’m not sure whether or not this was added in the revised edition of On Growth and Form (1942), but I’d like to think that this was written in the first edition, published in 1917: Read the rest of this entry »

If you had doubts before starting to read Neurdon, I think that by now you have come to the realization that there are as many neuroscientists that use computers than computer scientists that want to “use” neuroscience. This post is another example of the latter: Steve Furber, Professor of Computer Engineering at the University of Manchester and leader of the team that designed the the 32bit ARM microprocessor, is trying to build a large-scale neural simulator out of ARM processors. Read the rest of this entry »

The New York Times published an article this Monday on I.B.M.’s bid to have their Watson computer system compete in a number of televised Jeopardy! episodes, a move reminiscent of the famous set of chess matches between I.B.M.’s Deep Blue and Garry Kasparov. Reading the entire report may take some time but is definitely worthwhile for anyone who’s never heard a description of the set of problems involved in building such a computer system. Read the rest of this entry »

According tho this post, HP plans to introduce the first commercial product based on memristor memory in three years. In case you wonder: no, it won’t be a USB brain. According to Technology Review, it will be a flash memory.

Why flash memory? This storage suffer from some of the same limitations that plague silicon transistors: the limited amount of data-writing cycles, and the physical limits that prevent increasing storage in dense memory devices. Memristor memory can withstand up to about a million read-write cycles in lab tests, and can achieve densities unreachable by conventional technologies currently employed to build flash memory devices.

Want to learn more? Check out the original post.

In About SyNAPSE I characterized neuromorphic devices as the opposite of conventional Von Neumann processors. This is somewhat of a oversimplification, however. Modern processors are actually quite evolved from pure Von Neumann devices. They are dramatically more capable on virtually every computational workload than their heritage would suggest is possible. For neuromorphic devices to find any success in the marketplace, they’ll need to offer a significant performance gain against existing solutions, but with comparable or lesser power consumption and cost.

Read the rest of this entry »

Top Gun taught us that the best and brightest pilots can perform some amazing aerobatics.  Nobody seems surprised that a good pilot, with some practice, can move seamlessly from the flight maneuvers used on a Boeing 747 to those featured in Blue Angels shows.  While computer autopilots have performed well in commercial aircraft for some time, however, getting an electronic computer to pull a plane successfully through an aerobatic maneuver is almost impossible, and is thus a relatively new field of research. Read the rest of this entry »