Engineers find 'missing link' of electronics

[Indenturedebtor]

Source: New Scientist

"Nanoscale circuits based on molecules used in sunscreen lotion have led to the discovery of the "missing link" of electronics engineering – a previously mythical device known as a "memristor"."

******

"Chua showed that his predicted device could remember the last voltage applied to it, and how long it had been applied. He dubbed the property "memristance" but the memristor was quietly forgotten because it was unclear how it could ever be built."

******

"But Williams' team has now done just that, using nanoscale circuits including molecules of the active ingredient of sunscreen – titanium dioxide."

******

"Chua, now close to retirement, is thrilled at the finding. "This seminal work presents the first example of the memristor I postulated in 1971," he told New Scientist. "We can now expect many new unconventional applications, including super-dense memories and brain-like computing chips.""

--------------------------------------------------------------------------------------------



Read more: http://technology.newscientist.com/article/dn13812-engineers-find-missing-link-of-electronics.html?DCMP=ILC-hmts&nsref=news2_head_dn13812

---

WOOHOO :woohoo:

AI here we come! I know this is Science News but damn if it doesn't have huge implications for all humanity! This is the biggest development since the Transistor was invented!

[cosmik debris]

1. I'm flabbergasted!

Moore's Law wins another round.

[tridim]

3. Doesn't Moore's law deal only with transisters?

He'll have to redefine the law for memristers.

Thanks for posting this.

[cosmik debris]

15. Ooops! you're right.

My memory could use an upgrade too.

[Psephos]

77. Not transistors, but the increase over time in computing power per unit of cost

Moore's Law says that 18-24 months from now a dollar will buy you twice as much computing power as a dollar buys you today.

[Indenturedebtor]

4. There may have to be a whole new law

Moore's law only accounted for the first three tools of electronics... this is a whole nother dimension of computing possibilities. As a student of psychology I learned about the pros and cons of a neural type of processing vs an electronics approach.

Basically the brain as a computer is slower than a computer processor, but is able to perform millions of little calculations/programs at the same time... computers on the other hand can only do one at a time but they do them REALLY fast. By combining the two abilities to both process in serial and in parallel you are going to have ONE MEAN MUTHA of a computer! Seriously, I'm no electronics engineer but when this all gets hammered out your computer is going to be about 1,000,000 times faster pretty much overnight.

[Uncle Joe]

30. This could be Mary Tyler's law!

:think:

[customerserviceguy]

68. Naw

that just applies to the gravitational effect on a tosset beret...

[Winterblues]

87. Wow just think of all that porn that will be available

:shrug:

[roguevalley]

43. TIME OUT, SCIENCE DUDES AND DUDETTES!!!!! Tell me, is

what this means HUGE memory storage, etc?

[Indenturedebtor]

46. It's more than that

Tons of very fast memory at first... a little later neural networks!

http://en.wikipedia.org/wiki/Neuron

These have the potential to be arranged to function EXACTLY like neurons. Now besides being a parallel processor (multiple thoughts at the same time) the brain also uses varying levels of coordination such as chemical baths to "set the tone" for the mind... but such things could probably be modeled.

[hootinholler]

66. It's even more than that...

Once you have neurons, you have holistic memory, which will blow the minds of countless Info Retrieval geeks. The hardest part is going to be getting out of the notion of addressing memory, and into the notion of stimulating it. Jesus, I'm old, I worked with drum memory for christ's sake!

Hmmm, the scale is nano, I'm imaging a terabyte or two equivalence in a reasonable sized device.

-Hoot

[Canuckistanian]

61. Undoubtably, yes.

And with a thousand-fold increase in speed, if this can be implemented.

[Bake]

79. What the heck is Moore's Law? In layman's terms please

Bake

[slackmaster]

2. My memory seems to be super-dense today

Having a very hard time remembering peoples' names.

[Hydra]

5. And people call space the final frontier

We haven't even figured out everything on earth yet.

Correct me if I'm wrong, but doesn't this discovery have applications in cybernetics? Sign me up for wired reflexes, if so :evilgrin:

[Indenturedebtor]

18. Oh and yes it does have applications in Cybernetics

There are already chips that can communicate with neurons. Now if you have a chip that IS for all intents and purposes a neuron just think about the possibilities. Not only can you directly interface with the brain, but many of our movements are actually controlled at the local level. Most of the processing for walking for example actually takes place in the lower spine. Even if someone had say completely lost the lower half of their body, you could quite easily build them new robo legs that communicated and worked perfectly with the brain. Have a spinal injury? No problem, here's a robo neuron. Want to preserve your mind forever in a hot robo body? May be possible ;) Want a hot robo girlfriend? Why not :P Batteries not included though hehe.

At this point we can't even imagine all of the possibilities here!

Please excuse me while I step into my home Matrix.. I have a date with 30 hot girls in my sky palace ;)

[Hydra]

20. I think I want my improvements to be biotech

but being having reflexes that make you 3x faster than before is certainly worth considering...

I just had a bad thought. I bet the military had this already. Cybered up soldiers...:scared:

[annabanana]

35. I will settle for a third set of teeth to grow in to replace the ones

I've been abusing since I was 8 years old.

[pokercat999]

24. You mean I could actually have a decent golf swing before

I die?
On edit: forget it.....I'm 57

[piedmont]

6. Sorry if I don't cheer at the impending obsolescence of humanity. nt

[Indenturedebtor]

8. This is just another tool

Computers have allowed us to make MASSIVE strides in everything from medicine to farming to cosmetics. We use them for pretty much everything and they have allowed us to work less for the same results. You want to figure out how to harness Cold Fusion? You're gonna need a honkin computer. You want to model climate change? You're gonna need a honkin computer.

As with any technology there will be dangers and advantages. We just have to use it right.

[Bread and Circus]

9. Work less for the same results? And yet buying power for the average American is no better...

than 40 years ago. Living from paycheck to paycheck, less economically secure, more debt, and rising cost of non-discretionary expenses.

I'm a huge nerd and I love technology. I also realize the value of technology.

However, there is a disconnect between the ability to make all our fancy machines and the wisdom in how to use them to actually better people's lives.

[Indenturedebtor]

11. I agree

Instead our ability to produce more work and to take shortcuts at home using electronics has allowed our employers to squeeze more and more out of us. But the fault lies not in the machines but in our overlords. At this point in human development, without massively increasing our tech and using it wisely we are doomed. And so I celebrate this major advance!

And this will make it to market, this is a fundamental advance not a product!

[Bread and Circus]

12. It definitely sounds amazing and I look forward to it. Ultimately though I think our

biggest payoff in terms of overall life improvement and sustainability will be in the non fossil fuel energy sectors. And of course, better computing will only aid in that effort.

[Indenturedebtor]

52. Indeed and if you read the better article posted downthread

These things promise to be WAY more energy efficient than standard electronics!

[NashVegas]

22. Computers Aren't the Tool - WE Are

Techie utopians still haven't figured out that the only non-elite humans of use will be those who are necessary for running the machines.

See Bill Joy's (ex Sun computers) essay "Why the Future Doesn't Need Us."

[Dreamer Tatum]

75. Has greater implications for the obsolescence of certain fossil fuel uses nt

[On the Road]

7. Great News -- Thank You for Posting

So many of these things never seem to get to the market. But this is tantalizing. One of these years, there are going to be more major breakthroughs in speed and capacity.

[eShirl]

10. Feynman was right that there's plenty of room at the bottom

http://en.wikipedia.org/wiki/There's_Plenty_of_Room_at_the_Bottom

[Zhade]

13. Religion zero, science uncountable.

Woohoo!

[coriolis]

14. Oh, marvelous. Now Kirchoff's Laws are gonna be known as

Kirchoff's Maybes.

:shrug: ::hurts: :banghead: :dilemma:

[Canuckistanian]

41. This doesn't violate any of Kirchoff's Laws

Circuits will still have normally behaving currents and voltages.

The true worth of this discovery is that ONE device can replace a circuit containing HUNDREDS or THOUSANDS of devices.

[MilesColtrane]

16. When these new synaptic computers become self-aware and look around at the state of the world...

...and the monkeys running it, they won't stop vomiting.

[HuskiesHowls]

27. Have you read "I Robot" recently?? n/t

[BlueJazz]

31. Yep..We'll have to change the 3 laws around.....maybe add.....

..."Unless the person votes Republican" :) :)

[santamargarita]

17. Plant one in Bush's head, there is plenty of room!

K&R

[crimsonblue]

19. I wouldn't go that far....

All by itself, it has little practical use. But paired with new discoveries of graphene semiconductors and single-atom transistors, this will certainly keep Moore's law going for a while. Memristors also hold great promise for programmable nano-robots, which could do such handy things as float around in the body and chop up cancer or abnormal cells as they generate. We are only about 50 years from being able to make humans immortal. I see your :woohoo: and raise you a :toast:

[Greyskye]

21. As a semi-conductor professional...

...with 24 years in the semi-conductor memory business (EPROM, EEPROM, Bubble, SRAM, Flash (single bit and MLC), and PCM; I find this really exciting!

Just as a straight memory device, this could seriously revolutionize modern electronics, and quite literally herald a new information-age.

Exciting times!

[Zhade]

23. So, how long until I can get that mp3 implant in my brain?

:D

[Greyskye]

25. Depends.

Do you want socketed-wire implants, or do you want to hold out for the wireless version? They both involves brain surgery, but the initial tests on implanted Bluetooth devices showed an unfortunate causality for pre-cancerous tumor growth.

Let me know if you want in on the closed Beta tests. You can get in on the ground floor if you agree to those pesky waivers. :evilgrin:

p.s. You haven't really lived until you've hacked some poor fools wireless brain implant, and Rick-Rolled them remotely. :rofl:

[Indenturedebtor]

26. Lol

Yeah that bluetooth thing was no good. But these obstacles can be overcome.

[Zhade]

69. LOL! No digipathic hacking for me, thanks!

But I'm "never gonna give {it} up" - meaning the implant, of course. :p

[Canuckistanian]

44. As a Technologist with 20 years electronics experience, I agree

The potential for elminating cumbersome A/D (and possibly D/A) circuits with a single device is very exciting to me.

As well as revolutionizing memory strorage beyond mere binary constraints and speed limits.

If this this is at all workable on a large scale, it may be the single single biggest discovery since semiconductors and the PN junction effect.

Astounding.

[Duer 157099]

28. Translation for the electronically-impaired please?

First place I got lost is why symmetry predicted this - what are the relationships between the resistor, capacitor and inductor, and how does that predict the existence of this memristor?

And, any layman-level explanation of why this only works on nano scale?

I want to be excited about this, but I don't understand it yet! Help!

:hi:

[Indenturedebtor]

33. Here's the real question you want to ask:

"How does a neuron work?"

http://en.wikipedia.org/wiki/Neuron

----

The reason that this is so exciting is that it can actually sum up voltages before discharging on it's own. The neuron does exactly the same thing by taking inputs in from its various branches and then deciding whether or not to say something to other neurons that it's connected to. That is the system on which our brains are built. Millions of neurons together = brain = consciousness. Every single neuron is basically a tiny computer, or even a sum of tiny computers all making "decisions" on their own.

A normal computer truly only has 2 states = either 1 or 0. This new memristor has multiple states at each gate. It's the difference between having a vocabulary and being able to say 'yes' or 'no' really fast.

[Duer 157099]

40. OK, it's getting clearer now

And since symmetry has come into play here, then can you tell me what the biological/natural correlates are of the other elements?

[Indenturedebtor]

50. Well

There really aren't many. The neuron is capable of far greater computational power than a processor. A processor just says "yes" or "no" really fast and has zero memory so it can't learn.

When you build memory into a logic machine it can learn. That's the thing about neurons is that they learn and change over time, even at a basic cellular level. And further the brain can rewire itself so that neurons know where to send a signal when they get a specific input. It's that ability to make all these millions of little decisions at once, and to change how you make those decisions cell by cell that makes the brain so powerful and so "intelligent". Think of the neuronal decision making process this way - 5 of your friends tell you to vote bush, but you don't trust them very much, 3 of your friends tell you to vote kerry. Based on past experience you have decided to trust those three friends more, so you go and pull the lever for kerry. It really is that complicated on a cellular level, and shortly will be in computing!

http://en.wikipedia.org/wiki/Neuroplasticity

A normal processor on the other hand can only execute the commands that you give it. Now that this "cumulative input" step has been pretty much handled by this breakthrough, all you have to do is give the computer the ability to wire itself, give it a few hardwired goals that drive this sort of "living evolution" in the brain and you will have a new form of intelligent life in the universe!

[Duer 157099]

53. Analog...digital...no, analog!

Is that what this comes down to, full circle? We've gone from analog to digital and now we're excited at the prospect of analog again, lol.

I love the irony of that. What's old is new again eh?

OK, granted, it's new new because it's at a smaller scale than ever before and so has new applications, but still.

[Indenturedebtor]

54. But it's different

To get your laptop to do the same basic decision making process you need tens of thousands of lines of code, tons of hardware, tons of electricity, and a bunch of really clever programmers. This just does it on a molecular level! Imagine taking every computer in the world and shrinking it to fit in your pocket... and then somehow getting them to THINK without software.. oh and it uses very little power and is made from sunscreen lol.

Seriously... i'm so freakin excited about this. It's developments like this that we need to solve some of the worlds biggest problems.

[Duer 157099]

56. Well keep posting about it

I feel that it is indeed important but I'm still trying to grasp certain aspects, so keep posting because your excitement is infectious!

I'm guessing that Chua and Williams might be sharing a Nobel prize soon.

[Zhade]

71. Am I silly to think "Skynet"?

:P

[Canuckistanian]

48. That's a good explanation

We'd no longer be limited to the rigid structure of the binary system in order to get high speed computations.

Less steps, less controlling "software".

The possibilities COULD be endless.

[Zhade]

70. I think I just had a techgasm.

Oh, btw, I LOVE your sigline!

[Greyskye]

38. nano-scale

This article has a good explanation for your nano-scale question: http://www.nature.com/news/2008/080430/full/news.2008.789.html

Williams and his colleagues created a memristor while experimenting with very tiny circuits. They sandwiched a nanoscopic film of a semiconductor (titanium dioxide) between two slivers of metal (platinum). Those are standard materials; the trick is to make the component just 5 nanometres wide — about 10,000 times thinner than a human hair.

It's only at the nanoscale that the behaviour of memristors begins to be detectable, Williams says. Any larger and they behave just like ordinary resistors, where resistance is equal to the voltage divided by the current. Electronics were originally developed at a scale far too large to see these effects and only recently have researchers been able to work at that scale.

[Duer 157099]

39. Ah, thanks, I'll go read that now

I got the full Nature article too that I will (ahem) *try* to read, lol

[bemildred]

29. It's still just equivalent to a Turing machine. nt

[Canuckistanian]

45. We're talking about a MOLECULE here

NOT a whole complex computer running a sophisticated program.

Just imagine THOUSANDS of "Turing Machines" in a CHIP.

[bemildred]

47. A thousand microscopic Turing machines is still just equivalent to a Turing machine,

[Indenturedebtor]

55. Or a brain

But a really fast brain that doesn't get tired or cranky.

[bemildred]

59. I'm not sure about brains.

It has to be serializable to be equivalent to a turing machine. Or maybe it would be finitely-serializable, or something like that.

But a brain that did not get tired and cranky would be a big advance.

And it's still true that a very fast Turing-machine-equivalent is correspondingly better than a very slow one.

[Canuckistanian]

60. What is your point about Turing Machines?

First, let's see what a "Turing Machine" is. Wiki defines it thus:

Turing machines are extremely basic abstract symbol-manipulating devices which, despite their simplicity, can be adapted to simulate the logic of any computer algorithm (as we understand them). They were described in 1936 by Alan Turing. Though they were intended to be technically feasible, Turing machines were not meant to be a practical computing technology, but a thought experiment about the limits of mechanical computation; thus they were not actually constructed. Studying their abstract properties yields many insights into computer science and complexity theory.

So, a "Turing Machine" was a thought experiment designed back in 1936 which pretty much predicted the model for our modern computers and John von Neumann came up with ideas for it's physical design.

The basic premise is, give a machine almost any problem which requires complex logic or math it can eventually solve it and present you with a solution you can understand. Do we not have this capability now?

What more do you want from a machine? Self-correction? Self-learning? Self-modification of it's core instructions or actually redefining problems? Well, those things are already being implemented in a number of experiments.

The problem is, it takes massive amounts of SPEED and MEMORY to accomplish even the rudimentary qualities such that I've mentioned mostly, I think, because of the constraints of the arbitrary binary/Boolean "two-state" logic needed to accomplish it.

This new development could possibly enable problem-solving capabilities FAR beyond the imaginations of even Alan Turing.

[bemildred]

63. Well, there are definite theoretical limits to what you can do with a Turing machine.

One example:

http://en.wikipedia.org/wiki/Halting_problem

But I don't really have any point, it's just a subject that has fascinated me for a long time, particularly the relationship to Goedel's Incompleteness Theorem.

I'm not saying it's not a great discovery.

[Canuckistanian]

64. Well, you bring up a great point- what do we want from a machine?

Are they just sophisticated calculators or do they have the potential for real, intuitive, problem-solving intelligence?

Do we really NEED a "Data"-like robot from Star Trek who is treated as an "intellectual equal" or do we need autonomous machines that do all of our problem-solving work for us?

AND, like the Halting Problem, when does it quit or even know WHEN to quit?

[bemildred]

72. Well, the thing is, if they were really that smart, what would they need us for?

For that matter, it we got them to do all the work for us, what would we be for? Is it enough for the human race to sit around and screw and take drugs and amuse themselves for the next few billion years? What the Hell are all these people FOR?

[Canuckistanian]

73. You've got a point there..

Wait a minute, my machine is giving me an urgent message....

[Joanne98]

32. So does this mean Hewlett Packard has the patent?

First predicted in 1971, the memristor could help develop denser memory chips or even electronic circuits that mimic the synapses of the human brain, says Stan Williams who made the discovery with colleagues at Hewlett-Packard's lab in Palo Alto, California.

Should we load up on HP stock?

[ozone_man]

34. It could take a while to make a practical device.

Patents take about 18 months unless expedited.

I'll bet their notebooks are all dated, signed and witnessed, unlike when the transistor patent was awarded, not to the first inventor, but to the scientist who kept the best notebook (AT&T).

[Indenturedebtor]

36. Yes and yes

Though it won't pay off for a few years.

[Zhade]

74. So we'll be beholden to HP?

NT!

[Joanne98]

37. Here's another article...and a picture of it.

It's the tale of the lost circuit.

Thirty-seven years ago, Leon Chua, a professor at the University of California at Berkeley, mathematically theorized that scientific symmetry demands that there should be a fourth fundamental circuit element. Engineers were already familiar with resistors (which resist the flow of electricity), capacitors (which store electricity), and inductors (which resist changes to the flow of electrical current), which can be combined to build more complex devices. The fourth circuit, which Chua called a "memristor" for memory resistor, would register how much current had passed.

"He looked at fundamental circuit equations and noticed there was a hole," said Stan Williams, who heads up the Information and Quantum Systems lab at HP Labs, "There should be a device that remembers how much current flowed through a device."


An atomic force microscope image of a circuit with 17 memristors in a row. The memristor consists of two titanium dioxide layers connected to wires. When a current is applied to one, the resistance of the other changes. That change can be registered as data.

(Credit: J.J. Yang, HP Labs)Williams and other scientists at Hewlett-Packard are publishing a paper in Nature on Wednesday demonstrating that that these things actually exist. HP has a few discrete memristors as well as a silicon chip embedded with memristors. It's a first, according to HP.

If memristors can be commercialized, it could lead to very dense, energy-efficient memory chips. Scientists have made devices that function like memristors, but it took a good number of transistors and several capacitors, Williams said. Memristor chips would function like flash memory and retain data even after a computer is turned off, but require less silicon, consume less energy, and require fewer transistors.

A memristor effectively stores information because the level of its electrical resistance changes when current is applied. A typical resistor provides a stable level of resistance. By contrast, a memristor can have a high level of resistance, which can be interpreted as a computer as a "1" in data terms, and a low level can be interpreted as a "0." Thus, data can be recorded and rewritten by controlling current. In a sense, a memristor is a variable resistor that, through its resistance, reflects its own history, Williams said.

Varying resistance is the same principle at work with phase change memory. The difference in phase change memory, which will come to market later this year, is that changes in resistance are accomplished through a substantial amount of heating. A bit on a CD-like substrate is heated rapidly a few hundred degrees and then cooled. Depending on how rapidly the bit cools, the material becomes crystalline or amorphous. The different states--crystalline and amorphous--exhibit different states of resistance.

"We can get it (resistance changes) with less energy," Williams said. "It is a large amount of resistance change with a small amount of memory."

The secret sauce in HP's memristors is two layers of titanium oxide, a crystalline material consisting of one titanium atom and two oxygens, sandwiched between two metal wires. The bottom layer consists of standard, consistent titanium dioxide. The upper layer is missing a few oxygens--less than 1 percent--which creates voids. When a current is applied (via the wire) to the upper layer, the vacancies are pushed into the lower level of titanium dioxide. That changes the resistance of the lower level. Subsequent bursts of current can then reverse it.

"All we have to do is push around a very small number of vacancies in a crystalline material," Williams said. "We can switch it very fast, faster than we can measure."

Pushing the voids into the consistent layer of titanium dioxide does not change its characteristics otherwise. He likens it to bubbles in beer. "You can have bubbles in it, but it's still beer," he said.


Memristors in green. The wires in this image are 50 nanometers wide, which comes to about 150 atoms.

(Credit: J.J. Yang, HP Labs)Memory and storage are the new frontier for chip designers. The explosion of data will require new ways to retrieve and store it. Cloud computing? It's a big hard drive, if you think about it. Numonyx, the Intel and STMicroelectronics joint venture, is leading the effort to commercialize phase change memory. IBM is working on ways to store data through magnetic charges on a wire. Seagate Technology, Hitachi, Zettacore, Grandis, and others are working on different memory and storage concepts.

HP has largely exited the chip business, but it has increased efforts to license the intellectual property inside its labs. The company, for instance, will likely try to commercialize the crossbar latch technology, which allows molecular grids to perform calculations. (Williams also works on that.)

While memristors can be made on silicon chips, memristor devices will require engineers to learn a new circuit design discipline.

"The technology is in good shape. The big barrier is not whether you can make it," Williams said. "It is the effort to design new circuits."
http://www.news.com/8301-10784_3-9932054-7.html?tag=nefd.top

[BearSquirrel2]

42. Instant on ...

The immediate fallout from this will be flash ram on acid. Super fast memory that does not require memory to refresh and would likely get down to a similar cost as standard memory.

Basically, you would be able to turn on the computer and it would be on. No hibernation required. Though, it would likely require periodic rebooting just like any other computer.

I always hibernate my computer, and I think I only reboot about once a week. Typically, because of some update.

[Canuckistanian]

49. You're still thinking in conventional binary/semiconductor/software terms

New operating systems could exist with this technolgy that would do their own diagnostics several times a second, no "rebooting" or "hibernation" needed.

And with that kind of computing power and speed, they could even tailor themselves to YOUR way of using a computer - automatically.

Microsoft's stranglehold on exactly what you get for an operating system may be over.

[Indenturedebtor]

51. Yes they could evolve to your needs based on simple driving algorithms

Need a computer for a camera? give it a few commands, leave it over night, and SHAZAM it's now a perfect computer for a camera, the likes of which the world has never seen.

Need to know how to make cold fusion work? Feed it the problem and let it figure it out. Chain a few thousand of these together and come back in a week. Just make sure that you tell it it's good and don't let it manufacture anything on it's own :P

[Zhade]

76. Okay, I'm starting to get that "too good to be true" sinking feeling...

I understand this tech is real - but will we really see it? Will we LIVE to see it?

I'm a futurist, but also a pessimist, lol.

[Indenturedebtor]

86. I know what you mean

I've seen so many breakthrough inventions fall by the wayside. But as this is a fundamental building block with numerous capabilities I think you will see it very soon. The potential for gain is too great to let this one go.

[BearSquirrel2]

67. I said immediate ...

You're talking about things that have been hypothesized. It won't necessarily be true. Just look at what the AI guys have accomplished in the last 50 years ... virtually nothing. The best GO program can't even beat a good amateur.

Given the results that the fuzzy logic guys get with multi-state self correcting systems, I'm inclined to suspect that these predictions are correct and nueral emulating systems are the way to go. But I also suspect that it will require an incredible amount of work to make these new models work in a stable and reliable way.

[old mark]

84. In a few years...

...the computers we are all using now will by like B/W tube TV's are now.

mark

[Duer 157099]

57. But why does it only work at the nano scale?

Is this similar to how certain quantum effects are only seen in very small scale (quantum) systems?

I mean, now that they understand what it is and how it works, will they be able to scale it up, or is that not desirable because what they really want is to scale things down?

[Indenturedebtor]

80. I think it only works at the nano scale

Because the "device" is an individual molecule. Technically you could probably write some software for your computer that did the same thing: remembered how much voltage passes through a particular part of your machine, and then provide some sort of output when the voltage reached a certain level.

But if I understand it right, the molecule basically does this on it's own.

The exciting thing about this is that it works at such a small level. Pretty much all of the advances in computing equate to gettting smaller and smaller devices that work faster and produce less heat so that you can perform more calculations in the same amount of time. This does both and it will provide the ability to use new forms of processing.

Because of the size and the properties here, when the kinks are hammered out you will basically be stringing together millions of little tiny molecule sized processors. All that a processor does is process chains of "yes/no" commands at lightning speed. These do the same thing but the range between yes and no is MUCH bigger, and you don't need all the hardware that supports the processor to do it. Furthermore because the memristor is capable of existing in multiple states (voltages), you can have a cumulative input effect that determines whether or not an action is taken. So to just pull numbers out of the air: four "yes" of input = 1 "yes" of output, 3 "yes" = "no"! That is the basic framework on which our brains function!

[bronxiteforever]

58. Kick & R-

[MadHound]

62. And then there's the dark side to this

"Chua showed that his predicted device could remember the last voltage applied to it, and how long it had been applied. He dubbed the property "memristance" but the memristor was quietly forgotten because it was unclear how it could ever be built."

all the better to figure out what a person has been doing electronically

[Indenturedebtor]

82. Well

I'm sure that there are limits to how long it could remember the voltage though. Want to wipe your HD? Just supply some electrical whitenoise and it's gone.

[Jamastiene]

65. Yes, that means a girlfriend might be in

my future after all. :wow:

How will I get her a tan though if she's made out of sunscreen? Hmm...back to the laboratory. :P

This is astounding. Great news. :D

[Irreverend IX]

78. Will these memristors be easier to mass produce than today's microchips?

That would give them a serious advantage, but from what the article says I gather you have to isolate individual molecules for memristors to work well.

[TCJ70]

81. In other news: Skynet 1.0 launched. heh n/t

[LongTomH]

83. Molecular Nanotchnology is also coming closer.

I just posted this over in the Science Forum: Prototype nano assember brings molecular nanotechnology closer!

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=228x40015

[jmowreader]

85. One quick way this will make your computing life better

There are two kinds of RAM: static and dynamic. Dynamic RAM stores information as little charges, and the computer has to keep looking at every memory address on your machine to keep them charged up. This slows your memory down, which slows your computer down, and you need to dedicate space on your processor for a memory controller. Static RAM uses little switches to store the information. This is faster because they stay turned on all the time, but it's more powerhungry and it throws off more heat.

Now imagine a memory with the speed advantage of static RAM coupled with the power and heat efficiencies of dynamic RAM. Your computer would go like hell.

In fact, it might even be fast enough to run Vista at an acceptable pace.