Scientists Create First Memristor: Missing Fourth Electronic Circuit Element
Researchers at HP Labs have built the first working prototypes of an important new electronic component that may lead to instant-on PCs as well as analog computers that process information the way the human brain does.
The new component is called a memristor, or memory resistor. Up until today, the circuit element had only been described in a series of mathematical equations written by Leon Chua, who in 1971 was an engineering student studying non-linear circuits. Chua knew the circuit element should exist -- he even accurately outlined its properties and how it would work. Unfortunately, neither he nor the rest of the engineering community could come up with a physical manifestation that matched his mathematical expression.
Thirty-seven years later, a group of scientists from HP Labs has finally built real working memristors, thus adding a fourth basic circuit element to electrical circuit theory, one that will join the three better-known ones: the capacitor, resistor and the inductor.
Researchers believe the discovery will pave the way for instant-on PCs, more energy-efficient computers, and new analog computers that can process and associate information in a manner similar to that of the human brain.
According to R. Stanley Williams, one of four researchers at HP Labs' Information and Quantum Systems Lab who made the discovery, the most interesting characteristic of a memristor device is that it remembers the amount of charge that flows through it.
Indeed, Chua's original idea was that the resistance of a memristor would depend upon how much charge has gone through the device. In other words, you can flow the charge in one direction and the resistance will increase. If you push the charge in the opposite direction it will decrease. Put simply, the resistance of the devices at any point in time is a function of history of the device –- or how much charge went through it either forwards or backwards. That simple idea, now that it has been proven, will have profound effect on computing and computer science.
"Part of what's going to come out of this is something none of us can imagine yet," says Williams. "But what we can imagine in and of itself is actually pretty cool."
For one thing, Williams says these memristors can be used as either digital switches or to build a new breed of analog devices.
For the former, Williams says scientists can now think about fabricating a new type of non-volatile random access memory (RAM) – or memory chips that don't forget what power state they were in when a computer is shut off.
That's the big problem with DRAM today, he says. "When you turn the power off on your PC, the DRAM forgets what was there. So the next time you turn the power on you've got to sit there and wait while all of this stuff that you need to run your computer is loaded into the DRAM from the hard disk."
With non-volatile RAM, that process would be instantaneous and your PC would be in the same state as when you turned it off.
Scientists also envision building other types of circuits in which the memristor would be used as an analog device.
Indeed, Leon himself noted the similarity between his own predictions of the properties for a memristor and what was then known about synapses in the brain. One of his suggestions was that you could perhaps do some type of neuronal computing using memristors. HP Labs thinks that's actually a very good idea.
"Building an analog computer in which you don't use 1s and 0s and instead use essentially all shades of gray in between is one of the things we're already working on," says Williams. These computers could do the types of things that digital computers aren't very good at –- like making decisions, determining that one thing is larger than another, or even learning.
While a lot of researchers are currently trying to write a computer code that simulates brain function on a standard machine, they have to use huge machines with enormous processing power to simulate only tiny portions of the brain.
Williams and his team say they can now take a different approach: "Instead of writing a computer program to simulate a brain or simulate some brain function, we're actually looking to build some hardware based upon memristors that emulates brain-like functions," says Williams.
Such hardware could be used to improve things like facial recognition technology, and enable an appliance to essentially learn from experience, he says. In principle, this should also be thousands or millions of times more efficient than running a program on a digital computer.
The results of HP Labs teams findings will be published in a paper in today's edition of Nature. As far as when we might see memristors actually being used in actual commercial devices, Williams says the limitations are more business oriented than technological.
Ultimately, the problem is going to be related to the time and effort involved in designing a memristor circuit, he says. "The money invested in circuit design is actually much larger than building fabs. In fact, you can use any fab to make these things right now, but somebody also has to design the circuits and there's currently no memristor model. The key is going to be getting the necessary tools out into the community and finding a niche application for memristors. How long this will take is more of a business decision than a technological one."
Image: An atomic force microscope image of a simple circuit with 17 memristors lined up in a row. Each memristor has a bottom wire that contacts one side of the device and a top wire that contacts the opposite side. The devices act as 'memory resistors', with the resistance of each device depending on the amount of charge that has moved through each one. The wires in this image are 50 nm wide, or about 150 atoms in total width. Image courtesy of J. J. Yang, HP Labs.
Posted by: cyberbian | Apr 30, 2008 11:49:34 AM
HOLY SHIT IT'S SKYNET!
Seriously though, that is fascinating, it really makes you think how little we know about the computers, I believe Berners-Lee said today that the internet is still in it's infancy, it's news like this that makes me think so are computers! What we want is some quantum computers, more quantum information theory, along with memristors, and we'll be entering a new age. As cyberbian said, bring on the singularity!
Posted by: Tom | Apr 30, 2008 12:32:38 PM
Yes, the singularity is coming. This may be a "small" development that precipitates larger changes in the coming wave.
Posted by: palegray | Apr 30, 2008 12:35:39 PM
Cool. Are these guys going to create a spin-off company to commercialize the technology? Cyberdyne Corporation would be appropriate.
Posted by: allen | Apr 30, 2008 12:39:46 PM
A very awesome breakthrough.
Posted by: Sinistar | Apr 30, 2008 1:11:33 PM
PUT A RUSH ON THIS TECHNOLOGY!
OUR GOVERNMENT
BUSINESS
MILITARY
ARE ALL SO IN DESPERATE NEED OF A FUNCTIONAL WORKING BRAIN.
GO FOR IT!.
Posted by: tiredolewardog | Apr 30, 2008 1:13:05 PM
We the people of Oregon would like to be the first to welcome our new cyborg overlords. All hail Skynet!
Posted by: danarchy | Apr 30, 2008 1:18:51 PM
unfortunately, if it processes much like the human brain, there will be errors, much like the one in the first sentence/para.
"humain" should be "human".
sorry, that one really stuck out like a sore thumb. Other than that, good article.
Posted by: g | Apr 30, 2008 1:27:30 PM
Isn't that a very quick way for adding numbers with so little energy?
Posted by: null | Apr 30, 2008 1:36:41 PM
Humain brain... use firefox.
Posted by: oldarney | Apr 30, 2008 1:40:17 PM
Oh great! Does that mean that instead of simply stylishly displaying web pages from favourite magazines, my computer will act more human-like and now whinge and moan about every little grammar and spelling flaw that it parses as the page loads?
Posted by: Oh Great! | Apr 30, 2008 1:52:16 PM
Hurray for humains!
Posted by: Notspymag | Apr 30, 2008 2:15:49 PM
Great, so that digital logic final I just took is already out of date.
Posted by: Jonathan | Apr 30, 2008 2:34:40 PM
The Cylons were created by Man.
They Rebelled.
They Evolved.
There are many copies.
And they have a Plan.
Posted by: K | Apr 30, 2008 2:50:25 PM
One more thing i dont understand
Posted by: Dumbo | Apr 30, 2008 3:02:33 PM
Great Idea!!!! Where can I access the original paper written br Mr. Chua.
Posted by: Charles Manning | Apr 30, 2008 3:05:13 PM
I am so happy about this.
Posted by: Edward Woltin/Koin | Apr 30, 2008 3:14:54 PM
Actually the more human-like activity is to read that sentence and not even notice the error. Our brains are much more forgiving of errors than are digital computers. We can read and comprehend very poor spelling and very poor grammar. That's also why our brains are so much better at facial recognition. We can recognize similarities in 2 images of a face despite great variations in color, tone and angle.
Digital computers need a lot of code in order to calculate exceptions to a rule. They do a great job of following the rules, but they don't do a very good job of bending the rules.
Take music file sharing for instance. 80's style mix tapes were an acceptable form of copying music. Computers in this digital music era have no way of distinguishing and accepting that kind of rule bending.
Posted by: I'm Humain | Apr 30, 2008 3:15:44 PM
Kurzweil's estimation that 10^16 flops are required to simulate a human brain did not take this technology into account. Each memristor should be able to simulate a neuronal synapse much more efficiently, requiring far less operations to simulate a human brain. We just moved the singularity back in time a few years =D
Posted by: omtron | Apr 30, 2008 3:26:07 PM
This is definitely something that will pave the way for solid-state storage to become cheap. But the problem with many great ideas is that the business people think they are outlandish and ridiculous and somehow "impossible", as did happen with many of Nikola Tesla's inventions. I would really love to have my PC's memory retain itself after it shuts off! That would eliminate the need of a moving device for storage.
Posted by: Kenneth Clive | Apr 30, 2008 3:29:15 PM
Are there references for this article or is it a mangled recollection of stuff you've heard about? This article partially contradicts the Reuters article, which also has no references.
Posted by: Kevin McMurtrie | Apr 30, 2008 3:30:29 PM
and thus the world takes one more step closer to becoming ghost in the shell,
Posted by: Foss | Apr 30, 2008 3:44:56 PM
This is truely news! How often is there some completly new kind of electronic component?
Aside from all other aspects of life, I imagine this will have huge repurcussions in the gaming/entertainment industries. The AI aspects alone will be amazing. Musical instruments like the synth can take radical new directions.
Yeah, I missed the "humain" thing entirely lol. Informative article, though.
Posted by: Stevo | Apr 30, 2008 3:46:01 PM
How wonder how long the memristors retain information. Could they preform a partial recollection, as our brains do, in the event that charge memory begins to fade?
Posted by: Michael Payton | Apr 30, 2008 4:03:32 PM
If it is not instansiable in Silicon then it won't matter to me in my lifetime.
Posted by: dond | Apr 30, 2008 4:28:00 PM
i think this kind of breakthrough can take minimum of 8 years b4 it hit the market or mayb longer or shorter idk jus guessing
Posted by: x2zee123 homiiiii | Apr 30, 2008 4:28:33 PM
Here's the original IEEE Paper by Chua:
http://ieeexplore.ieee.org/iel5/8147/23413/01083337.pdf?isnumber=23413∏=STD&arnumber;=1083337&arnumber;=1083337&arSt;=+507&ared;=+519&arAuthor;=+Chua%2C+L.
Posted by: Roy | Apr 30, 2008 4:28:49 PM
Viz: Carver Mead's article in IEEE Spectrum ~1990 on NeuroMorphic Engineering for a detailed analysis of Analog VLSI vrs Digital VLSI in TeraScale switching circuits.
Posted by: NeuroMorphosis | Apr 30, 2008 4:35:25 PM
hmmm.
Off?
How, exactly, could you define an Off state given this technology?
Posted by: Glenn | Apr 30, 2008 4:48:56 PM
This is just what I need to finish the pattern buffer on my transporter. Where can I get a prototype?
Posted by: Bob | Apr 30, 2008 4:49:52 PM
This is gotta be huge. I think that maybe 30 years from now my grandchilden will wonder how we fared without them memristors, kinda like when the transistor was invented, or later, the microchip. I am sure this invention will revolution electronics in the years to come. And I don't think AI will be the most important application, but there will be myriads of new applications we haven't yet dreamt of.
Posted by: Luis | Apr 30, 2008 4:55:47 PM
It's going to be _hell_ if you're a programmer in 10 years :)
Posted by: xyr | Apr 30, 2008 5:03:39 PM
no more "does not compute"
no more "binary overflow"
Posted by: faiz | Apr 30, 2008 5:25:22 PM
the first time i learned that we use just "on" and "off" i thought, "how terribly inefficient", when there's a whole range of variable properties... (voltage, amperage)..
it'd be like restraining my midi controller to only sending 0 or 127.
didn't ever make sense.
good work boys.
Posted by: adam | Apr 30, 2008 5:37:20 PM
Cpmouetrs cnnaot raed tihs snteecne, but "humains" can. In the naer ftuure, hveower, tehy mhgit!
Posted by: PBKman | Apr 30, 2008 5:38:14 PM
This is going to make surfing porn hundreds of times faster! lolz @ internet
Posted by: Kiber | Apr 30, 2008 5:57:24 PM
Humain in the membrane
Humain in the brain!
Posted by: firezog | Apr 30, 2008 5:59:54 PM
Skynet exists, the UK military have recently launched part of the fifth incarnation of this communications network.
http://www.eadsdsuk.com/index.php?id=153
Posted by: Mik | Apr 30, 2008 6:07:27 PM
Just had a thought while posting that last comment when I got to the capcha page. Neural AI is going to make that bit kind of redundant.
Posted by: Mik | Apr 30, 2008 6:09:30 PM
Human intelligence works through prediction via invariant memories stored throughout the neo-cortex, not memories stored in any singular place. The neo-cortex is a collection of cells interlinked through literally trillions of connections, more connections than there are molecules in the known universe. This memristor is going to need to "remember" more than than resistance, and is going to have to be connected into an nexus, not a straight line. We have a long way to go.
Posted by: mike | Apr 30, 2008 6:18:38 PM
@adam:
Depending on the application, either "on" or "off" is really really efficient.
.
If the input to the IC responds to a certain threshold, and you can get away with either doing "off" = 0v and on {whatever the max voltage threshold may be for the IC, 3.3v 5v, 12v, 15v), building a voltage divider to get to the bare minimum is kind of stupid.
.
It's all about the application.
.
Besides, a lot of circuits are made of analog and digital.
Posted by: Some Dude | Apr 30, 2008 6:26:03 PM
I don't understand how this is seen as similar to the brain. The synapse is either on or off.
Posted by: ecor | Apr 30, 2008 7:32:14 PM
While the advancement of technology is a fascinating subject for many - myself included - I cannot help but shirk from the concept of a device able to emulate a human brain. I have contemplated the issue myself many times, and worked out the theory behind the solution, but the incredible power behind such a device should cause all to pause. A device which is able to learn and make self-conscious decisions could have unlimited power. Since it is a mechanical device, it can be "upgraded" in a way which humans cannot duplicate. And, assuming that this kind of device were to remain subservient to its possessor, whoever owns this device would, vicariously, have that same virtually unlimited power. As man has grown in power over his dominion, his potential for evil has likewise grown. The safety of mankind should be soberly considered as this branch of technology continues to develop.
Posted by: Peter | Apr 30, 2008 7:46:40 PM
ecor: the synapse sends electrical signals through action potentials, and it's threshold for discharging an action potential is variable (controlled by many factors, including how often and when the synapse has been "on"). By adjusting these thresholds, human memory is formed. As this article describes the memristor, it works similarly.
Posted by: omtron | Apr 30, 2008 8:29:35 PM
mike: True what you say. The memristor simulates the learning of a single synapse. To simulate the entire cortex, we would need billions of memristors wired in parallel. However, I don't see why this cannot be done.
Posted by: omtron | Apr 30, 2008 8:36:12 PM
@glenn: off = no energy source = no way to change its state
Posted by: on | Apr 30, 2008 8:37:07 PM
Dear humains, maybe it's just in Chile, but here intelligence follows life (sometimes...).
And physicist always believe that consciousness have to do with some physical phenomena, like Penrose and his fit-all-the-gaps-particles. The problem as I see it, it's the brain in a bat metaphor: people use to think in the brain as something apart from the body and the world in which it develops. Certainly you need a brain to think. And a childhood, and people to talk to, and time, etc.
Posted by: plaza | Apr 30, 2008 8:47:24 PM
As we advance to the '50s, 1950s that is, it is good to remember the devices that have brought us to this point. The Memister (no "r") is indeed an interesting concept when constructed using modern materials and techniques.
Posted by: Joseph E. Thomas | Apr 30, 2008 10:00:50 PM
GREAT! There will actually be much more interesting work out there in the future for us analog circuit design engineers (and more quantity of work). We just had to wait a REAL long time.
Posted by: Kman | Apr 30, 2008 10:43:13 PM
Learning machines will never become dangerous as long as they don't learn from us.
Posted by: Tiberian Fiend | May 1, 2008 12:12:00 AM
http://www.hpl.hp.com/news/2008/apr-jun/memristor.html
Posted by: sys | May 1, 2008 12:26:38 AM
Mike said"The neo-cortex is a collection of cells interlinked through literally trillions of connections, more connections than there are molecules in the known universe."
NOW, in works of 3 or less syllables, explain WHY that fdoesn't make any sense at all.
(hint...connections are made of more than 1 molecule)
Posted by: Lineswine | May 1, 2008 1:37:45 AM
I, for one, and I imagine Microsoft too, am alarmed by the prospect that (when turning your computer on) "your PC would be in the same state as when you turned it off." I thought the whole point of this procedure, was to get it going again.
Posted by: Richard Vahrman | May 1, 2008 1:45:57 AM
Richard, that is the funniest thing I have read here!
Lineswine, Mike probably meant (or should have meant) more POSSIBLE connections than there are atoms in the universe. Actual connections should be somewhat less...
The revolutionary thing here as I see it is not that we can now simulate synapses, analog circuits have been around for longer than digital and can already do that (BTW we don't need an A-D converter for each synapse we can build pure analog simulations), but that we could potentially build them with far fewer components, perhaps all passive. Its interesting that this completes the physical manifestation of some kind of mathematical truth, meaning we should find we can make circuits that are analogs of other physical systems in a simpler, more elegant way than before. Previously all sorts of weird functions were possible (e.g. a negative resistor) with active circuitry. Its the canonical passive aspect that makes this interesting. We used to try model any electrical or physical system using the three known components, and its actually very surprising to realise that in fact mathematically we were missing one and we chose not to use it as a modeling tool simply because we didn't know how to make one in real life.
Posted by: colo | May 1, 2008 2:57:54 AM
If we combine this with eestor's ultracapacitor, we'll have an electric car you can fuel up in 5 minutes that will drive itself for 400miles. See especially: http://bariumtitanate.blogspot.com
Posted by: jm | May 1, 2008 4:13:20 AM
My instructor was Mr. Langley, and he taught me to sing a song. If you'd like to hear it I can sing it for you.
Posted by: Rothschild | May 1, 2008 6:39:32 AM
As an academic electrotechnical engineer, I have to object to calling this a "fourth basic element". There is no "gap" for this component to fill. There only exist three _basic_ components, i.e. ones that rely on fundamental physical principles, and that have parallels in other fields of science (like capacitor/spring and inductor/mass).
I congratulate the researchers on creating an (apparent) efficienct implementation of a device that may be used extensively in analog or digital circuits; however, they have no ground for calling it a fourth elementary component.
Posted by: Charl | May 1, 2008 7:13:29 AM
Hmmm, high speed FPGA neural network cards... nice.
Posted by: Sanity | May 1, 2008 8:05:22 AM
Crazy shit i think they should start working more on that shit!!!!!!!!
Posted by: DIck boy | May 1, 2008 8:18:26 AM
"As man has grown in power over his dominion, his potential for evil has likewise grown. The safety of mankind should be soberly considered as this branch of technology continues to develop."
I want to thank Peter for articulating my thoughts on this matter better then I could have done myself.
I am not at all convinced that this is a good thing, in fact I believe it may be just the opposite.
Posted by: Mike B. | May 1, 2008 8:30:54 AM
In response to "Charl", there definitely is reason to calling it a "basic" element. You have the three "old" ones:
v=Ri
i=Cv'
v=Li'
and a "new" one:
v'=Mi'
When M is a constant, the "new" one acts exactly like a resistor. The "memristor" action really occurs when M is not a constant (e.g., when M depends on i'). That gives the device "hysteresis" in the i-v curve. Nearly any device that shows hysteresis in the i-v curve can be thought of as a kind of "memrestive" element.
Remember that springs and masses are not good analogs for capacitors and inductors, respectively. However, a spring-mass system is a good analogy for an LC circuit, where you can think of the inertia (the mass) as inductance and the spring stiffness as the capacitance.
Mechanical analogies for LC circuits work when you equate (for example) speed with voltage and force with current. This analogy works because voltage*current = power and speed*force = power. So, gear trains are impedance transformations, and transformers are gear trains.
If you need an analogy for a memristor, you need to think about *CHANGE* in speed being related to *CHANGE* in torque in such a way that depends on *CHANGE* in speed (or torque) and not absolute speed or torque. Any MECHANICAL system that displays HYSTERESIS (there are TONS) work this way.
Posted by: Ted | May 1, 2008 8:32:33 AM
This is sensationalist BS.
Everything ascribed to this device
can be done to whatever accuracy
desired, by a standard digital
computer; and it can be done today.
Your boss tell you you had to write
something that would get the clicks
and comments up?
Posted by: Cliff | May 1, 2008 8:40:19 AM
howyoudoingsuh?
Posted by: Amway Man | May 1, 2008 8:40:52 AM
Charl,
Thanks for the insight into the industry, I see your points, however the claim of the "Fourth Electronic Circuit Element" in my opinion is listed as missing as its absence has left us with undevelopped technology. We may not view it as missing now, but as a programmer, this type of news opens the floodgates to new ideas and how to interface with hardware. I'd reword it to say that we've opened an entirely new dimension in electronics which can be applied to computing to have profound effects.
Posted by: Jon No | May 1, 2008 9:01:11 AM
Charl,
How about a water line with a water filter, as the filter slowly becomes dirty overtime, the flow of water through it meets with more resistance.
Posted by: Dizzy Z | May 1, 2008 9:32:18 AM
Paging Doctor Forbin
Posted by: EightBit | May 1, 2008 10:24:57 AM
Digital is dead.
The Future(TM) is Analog.
Posted by: jim | May 1, 2008 10:31:04 AM
Great - with "instant on" technology, now even rebooting won't fix anything.
Posted by: Bozo the Engineer | May 1, 2008 11:34:50 AM
Does this mean my porn will just load right up when my wife turns the pc back on?
Posted by: uh oh | May 1, 2008 12:45:28 PM
@Jim
I think you missed the train back in the 70s. Analog died a long time ago and analog computers are a complete joke. (too much development time and not as versatile as digital)
This tech should help with the development of static ram that holds its value even when no voltage is applied.
Too bad the article is so dumbed-down you can't glean anything from how it works. When reading the voltage from the resistor, does it discharge like a capacitor does? I find it hard to believe that some device will continue to hold an electric potential when being scanned/observed and not revert back to some voltage-neutral state. Maybe I've played in the transistor world too long. And FYI, the original poster failed to list the semiconductor in his list of electronic circuit elements.
Posted by: Modal Realism | May 1, 2008 12:50:44 PM
... and you thought that there wasn't any practical/everyday use for chaos theory !!!
Posted by: isaac riley | May 1, 2008 1:32:46 PM
Cyberbrains, here they come.
Posted by: Guillermind | May 1, 2008 1:59:29 PM
Awesome stuff.
I hope they mean to model the higher functions of the human brain. Hollywood history shows that the downfall of HAL, WOPR, VGER, and SkyNet was directly related to the sci-fi scientists modelling the freaking reptile part of the brain along with the higher functions.
Posted by: Qwaz | May 1, 2008 2:37:05 PM
Ok, maybe I missed something, and if I have please fill me in here...
How do you measure a memistor without affecting the present resistive state? Keep in mind as the size is scaled down this issue becomes more important (smaller values to measure).
It seems to me you must affect the property you are trying to measure by measuring it, and I agree with the post by Modal Realism - over time isn't that value going to decay to a point where it no longer accurately represents the original value?
I suppose there could be some kind of 'read/re-imprint' process?
Thoughts?
Posted by: Bad Andy | May 1, 2008 4:57:56 PM
this makes four basic circuit elements??? what about the transistor, that's what drives computer logic??
Posted by: doogan | May 1, 2008 5:08:28 PM
yay singularity. Hurry up and unlock this technologies potential you stupid scientists.
Posted by: Ugg | May 1, 2008 5:29:34 PM
I'm curious to know... How long can the memristor remember its charge? Will the memory slowly leak, or is it more like solid state?
Posted by: geofflee | May 1, 2008 5:40:53 PM
It doesnt hold a charge It just holds the memory of whatever the resistance was when the power turned off, and its read with a.c. so it doesn't disrupt the charge as it reads it.
Posted by: dickalan | May 1, 2008 7:03:30 PM
How can you say bring on the singularity..
Then there is a great risk it will make you totally completely obsolete! In every sense of the world.
I do belive we are not yet ready for AIs or AGIs.
Posted by: mrr | May 2, 2008 7:01:12 AM
Singularity? RUN!!! My computer slows down and/or crashes everytime microsoft sends out updates. How would you like to have your brain do the same thing? Say, while landing a 777 full of people or during a calculus final? I just downloaded the "official" drivers for my printer from the manufacturers website and they show up in GREEK. They work, but they're not really very useful because I CAN' READ THE INSTRUCTIONS!!!
Yeah, Skynet is a scarry idea, but as long as we have Microsoft to install in it, it'll never work! All Hail Bill Gates! (Picture huge stadium rally decorated with 10 story banners with the Windows logo in the middle and 50,000 people chanting, chanting, chanting...Maybe it's in Nuremburg)
Posted by: mechanicphilosopher | May 2, 2008 8:07:41 AM
it is still a resistor what about a mem-cap or mem inductor
Posted by: will | May 2, 2008 11:59:27 AM
a transistor can be looked at as a varible resistor and is considered active and not passive like the others
Posted by: will | May 2, 2008 12:10:10 PM
Great, now when my PC crashes, and I have to reboot. It will restore my current state. I'll get stuck in an infinite loop :p
Posted by: Ed | May 2, 2008 4:28:38 PM
A circuit that remembers! It just might solve the age-old mystery of human consciousness. Let's see where it'll lead...
Posted by: fred | May 3, 2008 7:47:32 AM
@Qwaz, sir I do believe that you have mistaken the reptile brain for the Reptilian Complex - which is basically described as the part of the brain (specifically used in describing the human brain in correlation with the triune brain model) that higher mammals share with reptiles. It is responsible for rage, xenophobia, and basic survival fight-or-flight responses. Any attempt to model this into an artificial brain might be a bad idea, unless you're trying to model a business oriented thinking machine. That could be profitable.
Posted by: dr. r-tard | May 3, 2008 3:39:19 PM
Its a wonderful discovery. Congrats to the team !
Posted by: swapnil ghike | May 4, 2008 11:21:47 AM
I believe all nature of technology is to be commended providing it is tried and true and is beneficial to man.
Posted by: L.A. Goates | May 4, 2008 6:48:41 PM
I do understand this as a material closely akin to a memory material
Posted by: R.Gopalakrishnan | May 5, 2008 5:32:12 AM
1- No ideal circuit element exists physically; in other words all physical elements’ behaviour is valid only within a range of the values of physical variables like current, voltage, frequency, temperature etc. Apply 50000 V to a carbon resistor you will measure 0 A instead of 50000/R; by the same token no linear element exists physically.
2- However we use these basic elements (v=Ri, v=Li’, i=Cv’ etc.) as basic building blocks which are capable of modeling almost all existing physical devices; but unfortunately all these models are linear.
3- We sure need nonlinear basic elements as: i) the World is highly nonlinear, ii) nonlinearity is much needed in many applications (clippers, rectifiers, modulators etc.) It immediately follows that to model, calculate and put to use these nonlinear systems NONLINEAR BASIC BUILDING BLOCKS ARE NEEDED and Memristor is one of them.
4- An algebraic relation between charge and voltage defines a capacitor, between current and voltage a resistor, between current and flux an inductor. The missing link is an algebraic relation between charge and flux and it defines a MEMRISTOR. A simple memristive behavior is =M(q)q where i=q’ and v= ‘; if the memristor is linear then M(q)=M constant, and differentiation of both sides of the relation ends in v=Mi a simple resistor in other words. That is why in the linear world it is not possible to distinguish a memristor from a resistor.
5- Is memristor needed? Yes as a building block, for modeling and possibly for realizing complex phenomena at least with simpler topologies if it can be built simply and cheaply. In fact Leon O. Chua won a best paper award for his article on “Modeling the p-n junction with memristor” which I believe appeared in the International Journal of Circuit Theory and Applications.
6- The article’s author does injustice when stating “Unfortunately, neither he nor the rest of the engineering community could come up with a physical manifestation” as Chua also presented a 2-terminal circuit that behaves like a memristor but with the components and technology available then.
7- I would like to remind concerning Modal Realism’s “Analog died a long time ago” note that Digital exists because of our inability to store and operate on anything other than binary. Analog is much faster and is the future.
Posted by: Cem Goknar | May 6, 2008 3:36:05 AM
In response to mike, you see there are more connections in the brain than molecules in the known universe, could please explain me then what are those connections made of?
Posted by: Pobrecito hablador | May 6, 2008 9:12:13 AM
Its a wonderful invention.....it would be interesting how the circuit symbol of a memristor will look like...COngrats to the team
Posted by: Mukul | May 6, 2008 12:06:46 PM
In the dark ages, we used a bulb filled with mercury to tell the temperature. Then someone discovered the temperature-controlled resistor aka thermistor, so I could make a funky little digital thermometer for my kitchen (I'm intentionally ignoring thermocouples, beyond the scope here). We have voltage-variable capacitors and voltage-variable resistors, both of which make my linear circuits behave horribly. Take a piece of iron, it has memory of a magnetic field, we called it core memory back in the dark ages.
Now we have a charge-variable resistor. Way cool, but a 4th basic component? If its properties are the basis of defining a new component, then the previously mentioned ones also qualify, making it much higher in number than 4th.
Or, we can just call it a charge-variable resistor, say it's a wicked cool discovery or invention (not sure which), and move forward. Scientists come up with these relationships, now it's time for the engineering community to do something with them. Remember that guy that discovered that if a certain chemical is exposed to light it makes another material stick to the paper? What a useless concept, unless you're Mr. Xerox.
Maybe at the very least it will be another factor to encourage the great minds of tomorrow, presently considering college, to choose engineering, and expand this new technology.
Maybe the thermometer in my kitchen will now remember how cold it was yesterday and warn me to fix the problem...
Posted by: WayTooPractical | May 6, 2008 12:22:01 PM
I see a bright future of computers.
this rocking hardware will save a lot energy in research on dna-computers.
and like mukul i am looking forward for symbol of a memristor!!!
Posted by: Mohammed Saifuddin | May 6, 2008 12:47:53 PM
If folks are interested in more details, check out my blog. I answer some questions about the technology and provide links news articles with more technical information.
http://h20325.www2.hp.com/blogs/labsblog/archive/2008/05/07/6327.html
Jamie Beckett
Posted by: Jamie Beckett | May 7, 2008 3:06:43 PM
whoops! Wrong URL for more memristor details:
http://h20325.www2.hp.com/blogs/labsblog/archive/2008/05/06/6322.html
Posted by: Jamie Beckett | May 7, 2008 3:12:08 PM
I too question the concept of a fundamental "missing fourth element". The fundamental Maxwellian relationships between electron potential (voltage) and flux (current) are defined by the proportional, integral and differential functions, viz. voltage proportional to current (resistor); voltage proportional to the integral of current (capacitor) and voltage proportional to the derivative of current (inductor). Together these can be used to define an arbitrary linear circuit behavior using a system of second-order differential equations. There is no need for a fourth (or fifth or sixth...) fundamental element unless you need to describe a third- (or fourth- or fifth-) order differential system - in which case the additional elements would need to exhibit second (or higher) order derivative or integral behavior. The memristor does none of these things - it is a non-linear component (like many others) whose fundamental behavior cannot be described by a linear system of equations.
Posted by: Kevin | May 8, 2008 8:00:18 PM
My CD player will remember perfectly what I played the last saturday, but with no disc inside. Great.
Posted by: Pedro | May 9, 2008 9:32:46 AM
Well, the only problem I see with this is that if it is now an analog device that thinks, we no longer have control over it, in effect, we give validity to some of the fears driven by Hollywood movies where the AI takes over! Once the analog device is built, it cannot be reprogrammed without re-arranging the memristors physically!
Posted by: Supernatendo | May 11, 2008 3:55:57 PM
Bye bye DSP, top of the morning to Analog Signal Processors.
Posted by: Bubblesaurus | May 21, 2008 1:11:53 PM
I'm thinking digital camera sensors with incredible tonal and exposure range...
Posted by: BBB | May 21, 2008 2:56:29 PM
I don't think anyone ever did, nor ever will, say "darn, I can't make this circuit work because I don't have a Memristor!".
It doesn't do anything new that can't be done without one. If I needed one, I'd probably use a capacitor or a tunnel diode instead.
So I think this is hyped up and we'll never see circuits with more Memristors than capacitors and resistors.
FYI:Electronic neurons already exist.
Posted by: Dr. Mad Scientist | Jun 2, 2008 5:18:50 PM
do they have a electronic symbol for the memistor yet?
Posted by: nebworm | Jun 13, 2008 7:50:44 AM
You people need to settle down. The technology is real, its not any more troublesome than the invention of the cell phone.
Posted by: Andrew | Jun 27, 2008 10:29:26 PM
An electonic symbol image is included in the Wikipedia article (http://en.wikipedia.org/wiki/Memristor). I don't know how official it is.
Posted by: David Spector | Jul 11, 2008 3:39:38 AM
Memristor will surely help to invent new technology that seemed impossible
till date.It will and and modify concepts that will help us define newer technologies.
Posted by: josh | Jul 30, 2008 10:07:30 AM
It will be difficult to use the term 'computer' to describe advanced devices based on this technology. Also, transcending binary means that we must also transcend traditional code development techniques. It is possible that today's computers will still be used as an accessible gateway to the overly complex inner workings of the new machines. Performing endless cycles of teach/test/remediate for rudimentary knowledge, skills and abilities will be the only way to 'program' such a complex device to the point where it may be reliably used via a human-friendly interface. I assume that it may also be possible to clone fully indoctrinated machines on an assembly line so that each one need not be trained individually.
Posted by: Travis Lynch | Sep 11, 2008 8:43:28 AM
whoa, so there will be a time when computers wont take 20 years to load up?! now that would be awesome!
-jake
Posted by: Live Wire | Sep 17, 2008 1:48:27 PM
We should see a collective awareness emerge by 2013. It will be benign. Then the ultimate climax of technology that is the singularity will occur.
Posted by: Daniel Hazelton Waters | Nov 13, 2008 2:25:50 PM
Please correct me if I am wrong, because I am not a scientist or engineer... A byte in a 16-bit computer is made up of 16 switches (transistors) that are either on or off to represent one of 65536 possible combinations.
A Memristor could be used to replace all the bits in a byte, and depending on the accuracy of the measurements could replace any number of bits. So imagine, say a the equivalent of a 128 bit processor only having one component (a memristor) per byte.
The amount of components would be vastly reduced, simpler and less power consuming, so good news for the environment. You could program the processor to emulate any number of bits per byte (as I say depending on the accuracy of the measurement) without changing the number of components.
Like I said before I am not a scientist, but this sounds like good news all round.
Posted by: Derry Tift | Dec 15, 2008 10:32:28 PM
Memristors + cuantic computer = SIGULARITY( near of 2012 ).
Posted by: ESTEBAN | Dec 22, 2008 5:12:15 AM
@ palegrey (From the 3rd post): cyberdine the company, already exists. They make 'exoskeletons'. And their first product is called 'HAL' [Which if you remember, is the name of the crazy, murderous, computer from 2001: A space odyssey]. Robot -doom is closer than we think. You can look it op, its no joke. LOL
Posted by: Harshad | Dec 25, 2008 10:58:14 PM
I just found this article...Hope you like it.
Regards
Dawood
Posted by: Dr. Tilak | Dec 31, 2008 2:01:01 PM
check it out
Posted by: Dawood | Dec 31, 2008 2:02:03 PM
Chua was NOT a student back in 1971; he was a professor.
Posted by: Correction | Jan 8, 2009 5:19:40 PM
Chua was NOT a student back in 1971; he was a professor.
Posted by: Correction | Jan 8, 2009 5:20:43 PM
Chua was NOT a student back in 1971; he was a professor.
Posted by: Correction | Jan 8, 2009 5:20:48 PM
hey hey hey
Posted by: BOB SAGET | Feb 4, 2009 3:20:40 PM
your topic explanation is good ......
it would help me a lot if can mail me more detail about memresistors along with its internal operation diagrams,vi characteristics ....
thank you
Posted by: sandeep n bhat | Mar 6, 2009 8:44:31 PM
congrates to the team!
but i'm not clear about its working.
IS it as the charge flow through increase in memristor the titanium oxide layer increase the storage level in memristor sir?
is the flux produced by electric field is storing the charge...
I'm in need of details for my paper so please reply me soon ,sir!
Posted by: gowtham raja | Mar 30, 2009 6:13:40 AM
http://www.mensclothingstore.us/
http://www.nike-max-tn.com
Posted by: fgghh | Mar 31, 2009 5:27:02 AM
OMFG!
This is the equivalent of the perceptron tube in solid state.
Let the singularity commence!
This tube will enable analog neuron models to be able to store weighting values for individual synapses without using a Digital to Analog Convertor for each synapse. This means that Neural Network computers will no longer need to be bastardized to operate as discreet plains of neurons. They can be arranged to function asynchronously like normal neural configurations.
And many can be on a single chip. Not many DAC chips to support a single neuron.