6-Minute Memristor Guide

[rodhoffman]

I didn't watch the video yet but just to be fair I thought I'd put down that Leon Chua developed the theory behind memristors in 1971 (and 5th and 6th fundamental elements later on as a challenge to the industry many years later)

[karrek20]

Did u guys know that Nyle Steiner made a memristor out of Al and sulphur-treated copper or lead. Would it be more cost effective to construct your array out these materials?

[Ibringthetruth1]

Better than stupid wikipedia

[MrLesWhite]

where is this thing? Though it would be out by now!!

[pizdiceanu]

"very,very few electrical engineers..."?...in u point like,unidimensional brains(hihi).are u a mind reader?

[pizdiceanu]

plus,this "mmmmrstor" IS a galvanic element.so much with passive(agressive,homo league) behaviour.

[pizdiceanu]

identifying the integral of voltage vs time with magnetic flux is plain dumb.this fucko "mmmmrstor"(heavy nasal pronunciation,henry kissinger type) works in pure plan-parallel geometry:bye-bye magnetic connection.go do a self head shot in net video.

[pizdiceanu]

yet another washout impostor....

[spencertron88]

These oxidations kinda... wear off after a months... Not the best for long term data storage.

[Slayermbm02]

But can RRAM perform logic operations? I think that's what make this memristors revolutionary devices... I know HP still have to show some prof it does, but everything about it can't be just marketing, can it?

[Mniilla]

"whHHHhen"

[nlimchua]

Leon Chua, the Berkley prof who originally wrote the paper for memristors is from my hometown in the Philippines. He is the father of Amy Chua, Harvard law prof and author of the book, "Battle Cry of the Tiger Mom". Truly amazing that after 30 years since the paper was written about memristors that it is now in fruition. Not to mention, prof Chua is still alive to see it.

[skotters]

does this have anything to do with the whole graphene thing? I thought that was going to be the new architecture or whatever.

If you could compare, which would be better? (i'm no engineer)

[dmarkle1111]

@skotters : No, it has nothing to do with the use of graphene. You could probably make a memristor on graphene if you wanted, but fabrication would be a challenge since graphene is pretty bleeding-edge technology.

[coasterpro]

@ Kingsamuelson At the most fundamental level the memrister read scenario works exactly the same way as Flash; by tracking current flow. The difference is with Flash we're detecting the current between two gates resisted by a negatively charged oxide layer, and here we're detecting the current resisted by the TiO2-x/TiO2 ratio. The memristor read voltage is not enough to change this ratio, just as the Flash read voltage is not enough to drain electrons from the control gate. It's non-volatile.

[aadnk]

@Kingsamuelson: I believe they use AC current to read the state of the memristor, since AC would only cause the layers oscillate between two sizes of thickness, and not move the layer to either side as DC current.

[opus2k10]

wow my mind is blown. Those EE guys are geniuses

[Kingsamuelson]

Actually, its overhyped. The guy spent years of research trying to find some great discovery and didnt find much, but because hes human hes gonna overhype it. If you apply a voltage to change the resistance of one of these switches, well that is useless because when you want to READ the information there, ur going to need to apply a voltage to see whether its open or closed, meaning its absolutely useless!

[dennisjiewenliu]

It's not overhyped and it actually works. The fundamental properties of memristors were proven mathematically by Leon O. Chua forty years ago. As to its significance in computing, while you may not perceive a memristor crossbar doing the work of transistors strung together to physically perform logic, the idea Williams is currently discussing is a multi-layer hybrid chip with a transistor-type base addressing all the memristor layers. There's also photonic interconnect for future designs...

[zcold]

@noobenstein and @ drstupidistake why are you two arguing on youtube like a couple of retards? How about you contact this scientist and have the argument with him... fawk.. this thing is really interesting tho..

[fsalazar90]

@noobenstein This isn't moving protons around, it's just moving electrons. The voltage applied (or lack thereof) shifts up or down the fermi level of the wide bandgap semiconductor they're using (TiO2). This is what Mr Williams refers to when he says he is "expanding" the p-type region; the shifting of the fermi level down due to the applied potential essentially makes the p-type region of the TiO2 larger.

You mentioned knowing basic science. That seems to be as much as you will ever know.

[c433z]

Besides the monstrous task of changing how we develop software for this, there's the problem of titanium and platinum being much more expensive than silicon.

[tiftik]

telescope fail

[HHSWrstlr]

@noobenstein Sure memristors do exactly what transistors do...but they are passive, nonvolatile elements that use less energy and are approx. 1/15 the size of a transistor (3nm vs. 45nm). These are the future of electronics, they will provide us the technology to essentially combine the RAM and HD components of modern computers. As far as reliability goes, HP is in the process of creating a 20GB solid state device the size of a sugar cube...expand your research beyond YouTube

[cespedeslv]

Good information/Presentation, Mr. Williams..keep up the good! I'll be watching for new progress on "Memristor" with you and Mr. Leon O. Chua in mind.

[TCSC47]

But what I don't understand is how did Prof Williams overcome the retrogresive zimmer effect to make the thing work in the first place.

[drstupidistaken]

@noobenstein Actually, a transistor, the atom of the processor, is based on "moving parts" just as much as this memristor. Electronics rely on controlled electron movement.

[mingwk]

great... one new topic for EEE next sem....

[Atrus6]

@noobenstein There aren't moving parts though, the change is simply brought about by moving the electron deficenty. The memristor is like a combination of a transistor and a resistor, so you are able to do computations and store memory.

[Wwallace67]

I don't know if this man is telling the truth, or not, but breakthroughs are often made in a different manner than this man describes. Usually, some phenomenon is discovered, then applications are invented. The way this guy tells the story is surprising. They wanted to do X, and then invented Y to do X. Doesn't ring true.

Furthermore, I can't tell from this video how this discovery is significantly different from a crossbar utilizing semiconductor FETs, except, perhaps, in the physical size

[simanick89]

yeah.. this has about as much moving parts as a say a Electrolytic capacitor and we all know that those are bad idea... yo clearly don't know shit about how electric components work.

[08techgrad]

Stan Williams was working on the idea when some showed him Leon Chua's paper that illustrated the same idea back in 1971. Oh, noobenstien, much of your electronic components are chemistry related. The memristor will revolutionize electronic engineering int 21st century the same way transistors revolutionized amplifiers in the 1950s which soon lead to integrated circuits the we know today.

[Herooftimex]

@noobenstein you didnt discover it, he did. ill take the word of the scientist from HP as fact, not yours.

[bcg7752]

Excuse me, but Stan Williams of HP Labs did not invent Memristor or even develop the idea. Leon Chua of the Philippines did it. I have no doubt Prof. Chua will someday be credited for such invention and will receive a nobel prize for such accomplishment. Wait for a major breakthrough in the next few years that will change everything what we know.

[BrokenBjartur]

4:21-23 'whatwhenwmnmwhen we put a negative voltage...'

[TheNewHetfield]

hahaha

This discovery is interesting, but I think they are trying to sell it, more than showing it to the scientific community.

Anyway, if it's commercially possible, that will give a real boost to nanoelectronics, and a longer life to Moore's Law.

[greg642ster]

@BrokenBjartur A decrease in Voltage.

[sqhschief]

This is an awesome finding, I'm going to ask my professor about this.

[vodkacannon]

OMG game graphics will be like real life! we'll have 100x anti aliasing!!!!!!!

[deathcloud33]

I lol'd at the microscope thing.

[Xalimdor]

Great stuff. Looking forward to these things actually being used in a large scale.

[ILoveDogMeat]

WAITWHAT?

[zedpanda]

my prof was really excited about this... I'm in EE too and I'm actually really excited about this.. .

[WeaponXU]

I'm hoping to apply this knowledge to my master's thesis. This is some ingenious shit, fo rizzle.

[ramalama246]

I feel smart till I hear smart people talking.

[sckryde]

lol..

we're all smart..

we just need to learn material..

dedication and motivation doesn't come natural you need to practice and then it becomes second nature..

this is the basic fundamental of biological life..

exposure to an environment with specific rules will trigger the process adaptation matrix within the DNA.

so you see, just show up, and the rest, leave to your brain.

motivation and new abilities are simply by products of environment.

[deathcloud33]

"process adaption matrix within the DNA"? If you're going to make things up, at least TRY not to sound ridiculous.

[sckryde]

is there anything you would like to dispute?

or do you just wanna monkey talk ?

try diplomacy people might like you a little more..

[deathcloud33]

Try not sounding retarded and people might not think you're an idiot. I don't care if you like me; I don't even know you. I'm nice to the people that don't make idiots out of themselves.

[sckryde]

what are you talking about?

i think you're mixed up..

are disputing something?

what is this "like" business you speak of?

what exactly is retarded?

i don't remember insulting anyone or being rude. sorry pal i don't understand monkey talk..

[sckryde]

i doubt you could last against me in lab buddy..

im not even gonna waste my time with primates like you.

if you wanna debate a subject or explain your dispute.. that's more than welcomed.

but to start off with insults and comments backed by a perverse need you have to behave that way towards other people is unconscionable..

let me have it..

explain to me what makes your data correct and mine incorrect.

ill even lower my vernacular level so that we can communicate..

GO

[deathcloud33]

"exposure to an environment with specific rules will trigger the process adaptation matrix within the DNA."

There is no such thing as an "adaption matrix".

Trying to sound fancy doesn't really work when you seem to have forgotten how to use the shift key.

[deathcloud33]

adaptation*

[sckryde]

k thanks..

so you have no real argument..

i doubt you have any idea what the words even mean..

[deathcloud33]

Hrm, I replied but it didn't seem to go through so I'll repost it.

You do realize that you could VERY easily make a fool of me and 'win' by showing me ANYTHING that shows that there is such a thing as an "adaptation matrix"? And yet, you've done no such thing; instead, you've only thrown bad insults at me and tried to act superior.

[sckryde]

perhaps your comments are reflections of your own proclivities.

i don't remember insulting you.

in fact i don't even think i ever spoke to you.

[ramalama246]

Microscope. lol

[surfernase]

Well, if you are in the field, I'd call it pretty much common knowledge and nothing new Memristor = RRAM. Really nothing special here. Just one guy getting a hype giving it a new name

[agentbong138]

You are incorrect sir.

[MrLesWhite]

I need to do more research...good vid

[startgame316]

pretty interesting

[oscargrp]

Excellent video, thanks IEEE Spectrum for the information.

[tfburges]

Good stuff.