Inventor flames Reg, HP in memristor brouhaha Here is some Christmas holiday reading. HP's claim that RRAM, PCM, and MRAM are all memristor technologies is bullshit, the memristor is not a fourth fundamental circuit element, HP didn't find it, and its developing memristor product is a not a memristor - really. The flame from Blaise So says ex-USPTO patent examiner and …
...probably because it is easier that way to get media attention - by piggybacking HP. Now he is no longer some obscure person from organisation known for its, erm, thorough examination of software patent applications, but a scientist fighting for The Truth, basking in media spotlight.
"But why exactly the USPTO dude feel the Anger Raising In Him?"
Really, is it that hard to understand? HP claim to have developed a memristor; Chua's 1971 definition says a memristor is (among other things) linear; HP's device is non-linear.
You can interpret this as a broadening of the definition of memristor to include other memristive devices (HP's stance, which they justify by citing later work such as Kang's paper, so they have some support); or as an excessive claim enabled by unprofessional imprecision, which is Mouttet's position.
Many a flame war has been fought over less.
Personally, I think there's some danger in broadening the definition - like Robert E A Harvey above I'm concerned about losing the two-terminal requirement (hey, if it's a *fundamental* circuit element, it should be required to be fundamental, no?), and I think the passive requirement has to be kept too. I'm not so worried about linearity.
If you are going back to kill the HP board, can you do it at the time they took over Compaq.
While you are about it, can you go back to the time Compaq took over DEC.
Then Compaq's takeover of Tandem (thinking it was DEC...???wtf...???)_
Oh never mind.
HP is a dead duck these days so who really cares?
...but I frankly could not care less about whether the thing HP found is a memristor, or just a broader memristive system, or whatever. What I care about is whether it can do all the neat stuff HP claimed it can, stuff that was previously more or less impossible.
If it does, then by all means call it uzderopieöl or whatnot. The boffins can argue about whether it fulfills the definition well enough later, I want them in my computer NOW.
I'm very seriously concerned about them dropping 'two terminal' from the definition, because without that, then I can give you memristor behavior with a handful of transistors, diodes and resistors. It would, of course require an external DC supply - quite unlike what I would consider to be a real memristor.
...apply old snake oil!
With due deference to Stan Ovshinsky, who has been tilling this field before these young pups were born, can someone reveal the mechanical analogue of the memristor? Resistor, inductor, capacitor can be simulated with drag, mass, spring and the memristor is...what? This puffery reminds me of the ancient claims of a new mechanism: the Rolamite!
If anyone deserves credit in this field it would be Stan and Iris.
(BTW, I am a cranky old coot, and have always been one. When one accepts Thévenin's theorem one must wonder who this Norton guy was, and why he gained any street cred for stating the clearly obvious under the concept of duality. )
... that is what I was trying to envisage. To bracket the device it is somewhere between a Turing machine and Russian "Memory Water." As Arthur C. Clarke said, "Any sufficiently advanced technology is indistinguishable from magic."
(Let the peer-reviewed journals decide -- we're here for the tussle.)
Changes in current/rotational velocity are slow.
Flywheel definitely doesn't work for memrister because memristers don't operate by storing energy, while flywheels/inductors/capacitors do.
I'd go with the carpet analogy myself - a reversible, physical change in the media that affects the resistance.
Leon Chua's original definition was an interaction of charge and flux - more precisely a relationship between dphi/dt and dq/dt. The simplest physical mechanism I can think of is that the charge is a set of spins (i.e. charge) and their associated magnetic flux - a form of spintronics if you will. I don't particularly see it coming out of classical electrodynamics (Maxwell's equations) in the same way as a resistor, capacitor or inductor.
I was impressed with Professor Chue's teaching style in EECS 105, ended with my one and only A+ at Cal (this class was also my introduction to SPICE). He mixed in some real world examples with is discussion of the theory.
That diagram is not a classical resistor straight-line relationship as I was taught decades ago.
And yet it clearly should be as we are discussing the fundamental components, not non-linear resistors or capacitors with dielectric hysteresis or inductors with saturated cores and remanence as some apologists might imply.
Why is it drawn so? Any chance of getting an answer?
I agree that it looks very handy for HP to be able to get a patent on one device that is then stretched by definition changes to cover a whole class of devices.
As has been mentiond above, this is not about some guy feeling bitter about something. this is a guy worried that a patent held by HO for a very specific electrical device is being broadened to include almost amything that resembles it including, if taken too far, a powered circuit that pretends to be a memristor. This means that HP have a much better chance of claiming patents on technology extremely far removed from the original patent and stopping other people innovating. To me that would be a bad thing.
As far as I can tell, that argument is specious anyway.
HP et al should only be granted patents for specific devices they've constructed. If somebody else comes up with a different way to make a "memristor" then all power to them, and if the owner of an alternative patent complains they should be thrown out immediately.
Rather like the NPN transistor, IGBT and MOSFET are rather different devices with quite similar characteristics.
If he's worried about this becoming a patent land grab then it's the USPTO that are at fault.
Far as I'm concerned, the HP usage of a memristor is an electronics component, not a part of an old skool electrical circuit. But then there's no real reason why there should only be a missing "fourth component" to go alongside capacitor, inductor and resistor. You could add superconductor, diode and some other components into the list if you wanted.
Basically, it's the functionality that matters and not how you get there. What we need is something that works like a memristor as per the description of its voltage vs current graph but which can be easily manufactured in very large amounts in a very small space without any moving parts. It may be easy to build a memristor on a big hunk of circuit board using existing components, but that isn't going to revolutionise computing. If HP are achieving a memristor on an equivalent scale to the transistors found in modern chips then I don't care if they're called a memristors or a duckristors.
One of Blaise Mouttet's problems with memristor technology is the absense of magnetic flux in the devices found thus far, despite Chua having predicted the flux should be present.
This has not been peer-reviewed yet, but I have a draft paper up on arXiv (http://arxiv.org/abs/1106.3170) which identifies where the flux might be. A less turgid version has been written up for a journal and is currently going through the process of publication.
The main idea is that the charge in Chua's equation should be the oxygen vacancies (in the HP memristor) and thus that the magnetic flux should be that of the oxygen vacancies - which is an admittedly odd idea. However, using it makes the memristor announced by Stan Williams fit Chua's definition easily and the scales involved provides an explanation of why the magnetic flux has not been measured.
Secondly, my day job is making memristors and in doing that I have found that the ReRAM/RRAM papers very useful. As I can see it, the two fields are merging; top ReRAM people are editing memristor special issues in journals and memristor and ReRAM people are being invited to the same specialist conferences.
Dr. Ella Gale,
Unconventional Computing Group,
University of the West of England
If the oxygen vacancy leads to formation of unpaired electron spins, then there would be an associated magnetic flux. Since there are many materials where the resistance is dependent on magnetic flux, the presence or absence of unpaired spins could have an effect on resistance.
Chua's postulated interaction of charge and flux doesn't make too much sense in classical electrodynamics, but does make more sense at the quantum level.
I was going to give thanks to Mr Mouttet for using up 15 minutes of my life that I wouldn't get back any time soon, but in fact, I thoroughly enjoyed the discussion within the article, and indeed in the above comments. Didn't understand hardly any of it, A-S Level Electronics was oh-so-very long ago, but I still enjoyed it all.
Here's hoping it's a sign for 2012, more top drawer boffinry all round please, no more stooping to the lowest common denominator.
"as represented by Ohm's law; i=v/r"
Not a bad article, but it clearly shows that it has been written by someone with very little understanding of electronics and math. FYI: the international designator for voltage is 'U' (capital), not 'v'. Resistance is 'R' (again capital) and not 'r', and current is 'I (again capital); and not 'i'. So Ohm's law is 'I=U/R'
'i' is an imaginary number (i=SQRT(-1)) and becomes extremely important when looking at AC, capacities and impedances and the shifted relation between U and I, expressed through complex math.
Errors like this would have brought me a slap on my head in basic electronics training, and certainly shouldn't happen in an article that wants to discuss advanced electronics.
but I thought engineers used j as sqrt(-1) and mathematicians used i, and that mathematicians are wrong :)
Also, our lecturers used I for DC and i for AC. I can't remember if they generally used U for potential difference (though they nearly always talked about pd instead of voltage in diagrams)
Granted I got my lowest marks in my first electric circuits module and steered myself to computing modules, but I did slightly better in the communications and signals modules later on when there were fewer nasty matrices around (though less said about my comprehension of compound trig the better)
> the international designator for voltage is 'U' (capital)
Not when I was at University, it wasn't.
Do you have a reference for your claim? I remember a French attempt to force us all into using U rather than V, but I'm pretty sure it didn't get adopted - not universally, at any rate.
> 'i' is an imaginary number
In mathematics, that is so. In electronics, it is frequently used for AC current. The clash between meanings is resolved by using the term "j" for the imaginary unit.
> Errors like this would have brought me a slap on my head
You might yet get one...
Vic.
It seems to me the merits of this article are dubious:
Journalist writes article
Someone takes issue ('flames' allegedly)
Journalist writes rebuttal
The journalist should not have written the rebuttal. A knowledgeable 3rd party should have. As it stands, the rebuttal might represent a fine analysis but we can never know how objective the journalist has been drafting the rebuttal.
In particular the journalist has sided with Williams (which might be the correct position). However as a dispassionate reader the basic criticism (that the goal posts have been moved whether by Chua or Williams) doesn't really seem to have been addressed. Rather the journalist has just disagreed with the criticism (no surprise there). So we 3rd parties are no wiser.
Blaise Mouttet has sent in the following email, which is reproduced here:
Chris,
HAPPY NEW YEAR!
Regarding your recent article on the "memristor brouhaha":
Stan Williams is dead wrong that phase change memory (and other unipolar memory types) can be accurately considered a memristor (even given the recently revised 2011 definition which is arguably fraudulent in and of itself). Neither Stan Williams nor Leon Chua have any expertise in phase change memory (or various of the other memory types Williams claims to be "memristors"). No one as far as I know has provided a correct proof or reasonable model showing that phase change memory can be accurately modeled as a memristor or as a 1st order memristive system. It is irresponsible for Stan Williams to be spreading this type of misinformation for the consumption of the popular press unless such a proof is first produced and well vetted in the scientific literature. I believe that Williams is doing this for business reasons that have nothing to do with correct science. This is very dangerous if it is allowed to continue unchecked.
I am cc'ing David Wright who has a great deal of expertise in modeling phase change memory materials and who has published at least one paper I am aware of peripherally related to the memristor concept.
I am also cc'ing Dalibor Biolek who was one of the first researchers to create SPICE models for memristive systems and who recently published an important paper distinguishing between the pinched hysteresis loops of bipolar (TYPE 1, Fig.1a) and unipolar (TYPE 2, Fig. 1b) memory.
In the 1976 paper by Chua and Kang an example of an impossible Lissajous curve for memristive one-ports was given based on property 4 of memristive systems (see Fig. 5b). The characteristic of this curve is similar to that found due to the voltage snap back in phase change memory which David Wright or any other researcher in phase change memory can confirm. Thus either Chua and Williams are ignoring this property or (more likely) have too little knowledge of phase change memory to realize that a claim that phase change memory is a memristor is easily provable to be scientifically false.
The more recent article by Dalibor gets more directly to the point by noting in the conclusion that memristance-charge or memductance-flux relationships cannot generate pinched hysteresis loops of type-II (which are typically associated with unipolar switching found in phase change memory).
Willliams is making false claims about the memristor either out of his own ignorance or deliberate attempt at fraud to advance HP's business agenda (at the cost of legitimate companies developing ReRAM such as Sharp, Samsung, and Unity Semiconductor). This should not be allowed and serves only to create confusion in the public.
Given the above information Williams should at minimum withdraw his allegation that phase change memory is a memristor and apologize.
I am currently preparing an article which I will be posting on wordpress in a few days going into more detail about why various of the arguments made by Williams in his rebuttal are completely bogus.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
