Memristors will rock your world!

Message Bookmarked
Bookmark Removed

Anyone else fascinated by these poorly-named modern marvels? It looks like today, HP announced some progress, promising actual chips with memristors on them within three years.

If you haven't heard of 'em, here is some background info, and here is a bit more user-friendly article about the "missing link" in electronics.

schwantz, Wednesday, 26 November 2008 00:11 (seventeen years ago)

The Flux Capacitor cometh!

Nate Carson, Wednesday, 26 November 2008 09:42 (seventeen years ago)

yeah Wikipedia really is fucking garbage for trying to explain things in simple terms.

Ant Attack.. (Ste), Wednesday, 26 November 2008 10:02 (seventeen years ago)

yeah Wikipedia really is fucking garbage for trying to explain <s>things</s> science <s>in simple terms</s> to people who don't know jack shit about science.

That's about as clear an article that you're going to get on the subject, if you don't know what flux is or how to parse the most basic of mathematical equations then you just might have to settle for understanding a bit less than you'd like to.

NoTimeBeforeTime, Wednesday, 26 November 2008 17:19 (seventeen years ago)

FUCK YOU DICK HEAD

Ant Attack.. (Ste), Wednesday, 26 November 2008 17:43 (seventeen years ago)

Maybe someone should explain the formatting help to NTBT in simpler terms.

One Community Service Mummy, hold the Straightedge Merman (Laurel), Wednesday, 26 November 2008 17:45 (seventeen years ago)

Here is an even more easy-to-wrap-your-head-around article about the subject. The part I didn't get at first was how you would read the "memristance" values without altering them. Someone pointed out to me that you would write with DC, read with AC. Cool stuff.

schwantz, Wednesday, 26 November 2008 17:48 (seventeen years ago)

The part I didn't get at first was how you would read the "memristance" values without altering them.

That is explained near the beginning of the wikipedia article, in the "memristor theory" section. You've basically proven the point I made in my last post, thanks.

NoTimeBeforeTime, Wednesday, 26 November 2008 18:56 (seventeen years ago)

First of all, you are a dick head. Second of all, when I had this question THERE WAS NO WIKIPEDIA ARTICLE WRITTEN ON THE SUBJECT. I'm an electrical engineer, so I was interested in the breakthrough.

Sheesh. Talk about angry nerd dork.

schwantz, Wednesday, 26 November 2008 19:20 (seventeen years ago)

Can someone tell me how it is that these things become more efficient as they get smaller?

Nate Carson, Wednesday, 26 November 2008 19:34 (seventeen years ago)

Can someone tell me why these things are important. I know "jack shit about science" but I'd still like to know. Thanks.

Holden McGroin (Ned Trifle II), Wednesday, 26 November 2008 20:04 (seventeen years ago)

It might be less that they become more efficient as they become smaller, but more that they become more efficient AS SWITCHES (like semiconductors) as they become smaller, since there are less ions to move.

They are important for a few reasons:

1. You can create new types of memories that retain their state without power (and generally in a smaller size than a typical flash memory cell).

2. It is a new building block for analog circuits that could have applications in, for example, neural networks. Up until now, every "neuron" in a neural network consists of a set of inputs, and then some small piece of logic and memory that stores the weights for each of the inputs, samples them, figures out whether to trigger the output, and adjusts the weighting coefficients based on feedback from some other piece of logic. Using memristors, you could create a more natural model for neurons which would do all this without the need for logic and memory cells.

The downside is that for now they are slow, at least when used as switches.

schwantz, Wednesday, 26 November 2008 21:20 (seventeen years ago)

You will be able to have more, faster, and better pr0n, NTII.

sheepie (libcrypt), Wednesday, 26 November 2008 23:50 (seventeen years ago)

Thanks Schwantz. Here's the line from the article:

"This new circuit element solves many problems with circuitry today--since it improves in performance as you scale it down to smaller and smaller sizes," said Chua."

Would you say "improved performance" is analogous to being "more efficient as switches"?

Nate Carson, Thursday, 27 November 2008 01:28 (seventeen years ago)

It seems like, from the context:

Normal circuits, a they scale down, need to be refreshed more and more to hold the tiny amount of charge that distinguishes a one from a zero in a memory cell. With this, you would get two benefits:

1. As they scale down, he seems to be hinting that the switching time improves.
2. And, since they don't need to be refreshed, you don't waste the power that you do with normal memory.

But maybe I'm missing something.

schwantz, Thursday, 27 November 2008 03:36 (seventeen years ago)

I guaranty you're missing less than I am. But it's interesting anyway.

Nate Carson, Thursday, 27 November 2008 07:13 (seventeen years ago)

Here's a new article on the subject by one of the guys at HP:

http://www.spectrum.ieee.org/dec08/7024

schwantz, Monday, 8 December 2008 06:06 (seventeen years ago)

Awesome. Think this has affected HP stock prices yet? Might it?

Nate Carson, Monday, 8 December 2008 09:55 (seventeen years ago)


You must be logged in to post. Please either login here, or if you are not registered, you may register here.