HP Invents “Memristor” Element that Can Enable Energy-Independent Memory

[3v0]

I agree with Nigel, it might start a revolution in electrical engineering but the chances are it won't.

Yup, just like that LED gizmo that never caught on, and that dud know as the ink jet printer.

Sorry I could not resist. And for the most part I agree. But it is HP Labs and they have come up with their share of things that worked out.

In honor of HP Labs 40th anniversary, here are 40 innovations with impact
40 innovations that counted

 

[Nigel Goodwin]

Interesting, I know of no other passive two terminal device announced in a similar fashion over the decades. If you would Nigel, point to one?

I think you mean devices in general, such as non linear semiconductor devices, ICs, new IC technologies, new materials etc.

Exactly any devices - but the announcement didn't say they could make them, of if it was even possible - it looked more to me like a very hopeful announcement in order to try and attract funding.

A bit like the wireless power announcement from MIT last year or so

 

[Leftyretro]

Exactly any devices - but the announcement didn't say they could make them, of if it was even possible - it looked more to me like a very hopeful announcement in order to try and attract funding.

A bit like the wireless power announcement from MIT last year or so

I googled around a few days ago on the new HP development and it seems they have developed cross-bar functions with this new tech.

This could be a very big "Next Big Thing". Or not

Lefty

 

[3v0]

If at all possible HP will get the device to market because there are gobs of money to be made. It is not a pie-in-the-sky university project looking for funding.

Exactly any devices - but the announcement didn't say they could make them, of if it was even possible - it looked more to me like a very hopeful announcement in order to try and attract funding.

A bit like the wireless power announcement from MIT last year or so

Not so.

HP spends 3.5 Billion US dollars each year on research. Of that HP Labs gets about 8% or $280,000,000 dollar. When HP Labs wants basic research done in a specific area it often contributes money to universities as with the Trimaran project.

**broken link removed**

The work at the labs is at an intermediate level between the product research within other parts of HP and the basic research at universities and national labs. Such high-tech research, which is often far ahead of actual shipping products, can take some unusual directions.

 

[Sceadwian]

Nigel the device isn't a theory, they've made them. Here's a link to the article I orginally read, even has a atomic force microscope image of the device. The last paragraph pretty much says it all, any fab plant out there can make the devices, but the trouble is going to be getting circuits designed that use them and finding a niche market for them where they're better to use than existing technology. The problem with it being a fundamentally new device is there are no existing design rules to start with, so the initial circuit development and testing is steep hill.

https://blog.wired.com/gadgets/2008/04/scientists-prov.html

 

[killivolt]

I am thinking that you might begin to think Linearly. What I mean is that super computer's remember logic and process it as algorithms achieved. Once that is achieved begin a new process.

The Data storage and re-writing with this Memristor will allow computational software that may begin to run independently by starting several processes at a time and once one of the processes has located the desired output. It may begin again very rapidly from that point.

Maybe deploying FPGA fields or many FPGA's with a network of these memristors you can write data and re-write data at speeds currently unknown today.

This done with one computer and one processor. It will not need to do anything except observe anomaly's and run sub routines associated with data.
The inputs and outputs are adding or subtracting the amount of FPGA's which are re-written continually. Until the desired output is achieved or it has run all conclusions ?

 

[xaviergisz]

The articles about memristors mention that an active equivalent of a memristor can be made. Can anyone post a link to this kind of circuit?

 

[MrAl]

Hello,

I constructed a 'memdiode' one time in the past. It started out as a diode
with about 0.7 volt drop and ended up as a resistor with infinite resistance.
This happened just by turning the TV set on one day and blowing one
of the low voltage rectifier diodes

On the more serious side, i have to think about what Nigel said too because
after reading about superconductors changing America back in the 80's i
still do not own one device that has a superconductor in it. It's all in the
big labs and the LHC and stuff like that.

If it does work out in the long run and we actually see some of these
devices, it will be really great, except gee i'll have to come up with
another Spice model somehow...probably with an integrator and FET
or something like that. Who says it can't be built with other circuit
elements? Not me. We can build much more complex devices out
of other more basic circuit elements so there is no reason why we
can not construct that too. Heck, it should be much more simple
than some magnetic models.
Still, it being a single device controllable with a simple current is
very nice. I can see the apps coming if it all works out...but i
also dont see any repeatability data...how well do the properties
repeat? If this works out too, then one app would be a very
simple battery state monitor (battery gas gauge), to monitor the
state of charge of a typical cell. But then, how much energy
will it require to run. How long before the chemical activity
starts to diminish.
There may still be some hurdles to get over before we see anything
reaching the production lines.

 

[killivolt]

The most recent artical 4/8/ 2010 Memristor

Apology's now but, I couldn't resist it.

 

[Hero999]

That's already been already posted in another thread.

 

[killivolt]

That's already been already posted in another thread.

Crap I looked at some but didn't check the date ?

Oh, well.

 

[Hero999]

https://www.electro-tech-online.com/threads/memory-of-the-future.105900/

 

[Voltz]

Well... the Theory has been in the pipeline for years ... it's potential is great but it's old news When they apply it in a computer or other electronic application THEN I'll be interested - BBC Article - BBC NEWS | Technology | Electronics' 'missing link' found
And I don't see what's wrong with the sign

 

[Hero999]

Did you click on the link I posted previously, then click on the link on the first reply in the thread?

For some reason every time there's a slight improvement on memristor technology, the forums become unintentionally spammed with it. Old threads get bumped and people go posting the same links to articles.

 

[MrAl]

Hi,


The other thread seems to have been changed or is there another thread too?
There appears to be no info there anymore so this is the thread again i guess?

 

[Hero999]

They removed it, as soon as they discovered this thread.

 

[killivolt]

https://phys.org/news/2016-02-memristors-digital-memory.html

kv

 

[MrAl]

Hi,

Wow, old thread i forgot was even here. Good information and nice link still though.

So they got a little farther in 8 years. Whoopie, now just another 100 years and we will actually have one
I cant wait so i hope it is a lot sooner than that.

Think about it...
Memories that do not need power.
Potentiometers that hold their setting after being 'adjusted' with an external voltage.
"Relays" that have variable resistance rather than just closed/open functionality.
I suppose this will open the door for a new class of transistor too, the memtransistor (note MOSFETs are almost memtransistors already)
And memmem's, or mem^2, devices that remember what they once remembered (ha ha).
Geeze, memtransistors could result in entire new classes of stuff like memHbridges, transistor H bridges that keep their 'on' or 'off' state or anything in between.
See what you can come up with.

Now all we need are meminductors and memcapacitors

 

[nsaspook]

I think HP used one in 1939 with their first product. https://www.hp.com/hpinfo/abouthp/histnfacts/museum/earlyinstruments/0002/

A light bulb.

The Model 200A began as the subject of Bill Hewlett's master's thesis at Stanford University in the late 1930s. Bill had the innovative, elegant and practical idea of using a light bulb in a Wein bridge oscillator circuit to solve the problem of how to regulate the output of the circuit without causing distortion. Other oscillators that were available at that time were costly and unstable. By the clever use of the light bulb, Bill was able to simplify the circuit, improve the oscillator's performance and reduce the price.

It only remembered for a fraction of a second but you have to start somewhere.

 

[MrAl]

I think HP used one in 1939 with their first product. https://www.hp.com/hpinfo/abouthp/histnfacts/museum/earlyinstruments/0002/

A light bulb.


It only remembered for a fraction of a second but you have to start somewhere.

Hi,

I'd call that a "tempmemristor" because it can only remember settings from a very short time ago and forgets easily. Much like old age