I am apprehensive to post on the topic as I would like to limit my posts to more random musings of electronics than any commentary or repeat of hot-topics in tech news. I remember when I first learned of the memristor – it was this article in IEEE Spectrum. It was based on HP’s research paper in the journal Nature.

Side note: The same summer, I was working at Cypress Semiconductors in Lynnwood, WA, as an Applications Engineering Co-op. One of the engineers, Mark Lee, was a PhD in electrolytics and an otherwise awesome guy.  He entertained my fascination with the memristor accompanied by a couple diagrams. It’s a fond memory. End side note.

This blog in Spectrum by Dexter Johnson has a video which has been, for me, the best explanation of how hp’s device works. Following his link to a blog by Frogheart illuminates more more fascinating research going on with memristors.

On a different note, this ARS article about a Science publication by a couple of IBM researchers discusses possible limitations of current semiconductor technology. The article, titled Moving beyond silicon to break the MegaHertz barrier, is more about the speed limitations – as I understand it – due to switching times for voltage levels as currently required by fets. The article also is bout the general limitations of current semiconductor technology.

Now, I might be reaching a bit too hard here, but maybe memristors can help out. In the previously mentioned video, Stanley Williams states that “we should theoretically be able to reach into a certain integrated circuit, rip out 10 transistors and replace them with one resistor” (this is the last thing he says). Not sure how HP’s will help with speed, but it’ll definitely help with size.

This NY Times article covers an interview with Williams:

The most advanced transistor technology today is based on minimum feature sizes of 30 to 40 nanometers — by contrast a biological virus is typically about 100 nanometers — and Dr. Williams said that H.P. now has working 3-nanometer memristors that can switch on and off in about a nanosecond, or a billionth of a second.

This is with two years of development on the Titanium dioxide Memristor

Sounds promising.

UPDATE April 15th:

Here’s one more article on IEEE Spectrum to add to the stack. The article clarifies that memristors can’t replace the NAND gates of silicon logic, but “wants to perform something called the material implication, or IMP, function, says Williams. Logically, IMP can be thought of as p implies q, or if p then q.” According to the article, this can still do any boolean function.

Williams acknowledges that memristors won’t completely supplant silicon logic gates. Because memristors can’t inject energy into a circuit, silicon transistors are needed to drive them. The good news, he says, is that a single operation in a silicon transistor can trigger computation in multiple memristors. He notes that a processor featuring a grid of memristors that operates parallel to a grid of silicon transistors might be two or three times as large as it would be if it only had the silicon. But because the number of simultaneous calculations achieved by the memristors is the square of the number of transistors, tripling a 1000-transistor chip’s size by adding memristors would yield a thousandfold improvement in computing power with a negligible increase in power drawn.

Side note: the logic gate and truth table for IMP function:


Advertisements