I have a MS in EE and Computer Science but ...

[EarthwormJim]

Originally posted by: Train
I wouldnt say it's not getting enough media. The NYT's had an article on it the other day. Depending on what news headlines you keep up with ( I have a pretty tech heavy google home page) I have been seeing a TON of articles on this throughout the week.

While they always say it will "revolutoinize" the computer industry, remember we've seen a lot of promises like this before, some make it to market a lot less fruitful than thought, others never make it at all. Remember the holographic storage that was supposed to give us 5TB, super fast, super cheap, drives? Well I remember reading in 2000 that by 2005, they would be the norm. I have yet to see a commercial holographic storage device.

Though I have to admit this looks a lot more promising, HP is the main player behind it, and commercial uses look to be pretty close. While I dont think it will reach its full potential for quite some time, powerless memory seems to be a short term gain using this tech.

This is a little different though, since it's a passive circuit component. It's not something that has an isolated use, like holographic storage. This has ramifications for ALL of electrical engineering.

Think about how revolutionary the capacitor, inductor and resistor are. They ARE circuitry. Think about all the things we can make with just those 3 simple components. Now add in a possible fourth.

 

[George P Burdell]

Sweet mother of Ohm, how will we perform RLC circuit analysis now?

Did anyone read the paper and understand how it is supposed to work in theory? Wikipedia says that at a given instant of time, it acts like a linear resistor; but the value of the resistor M changes with time depending on charge. So how exactly would one change M?

 

[ranmaniac]

I'm looking forward to DNA Computing:

http://www.peterindia.net/DNAComputing.html

Computing with DNA offers the advantage of massive degrees of miniaturization and parallelism over conventional silicon-based machines. For example, a square centimeter of silicon can currently support around a million transistors, whereas current manipulation techniques can handle to the order of 10^20 strands of DNA.

Size: the information density could go up to 1 bit/nm3.
High parallelism: every molecule could act as a small processor on nano-scale and the number of such processors per volume would be potentially enormous. In an in vitro assay we could handle easily with about 10^18 processors working in parallel.
Speed: although the elementary operations(electrophoretic separation, ligation, PCR-amplifications) would be slow compared to electronic computers, their parallelism would strongly prevail, so that in certain models the number of operations per second could be of order 10^18 operations per second, which is at least 100,000 times faster than the fastest supercomputers existing today.
Energy efficiency: 10^19 operations per Joule. This is about a billion times more energy efficient than today's electronic devices.


I hope Crysis can run decently on this.

 

[Minjin]

It was in today's (Thursday) Wall Street Journal so the media is definitely covering it.

 

[Qacer]

From the article,

Originally posted by: Analog
the inductor, which turns current into a magnetic field.

Somehow, I'm not at ease with this statement. Ampere's law relates current on a very long straight wire to a magnetic field. I don't see how an inductor "turns current into a magnetic field", when current and magnetic fields are mutually related already.

 

[Analog]

Originally posted by: Qacer
From the article,

Originally posted by: Analog
the inductor, which turns current into a magnetic field.

Somehow, I'm not at ease with this statement. Ampere's law relates current on a very long straight wire to a magnetic field. I don't see how an inductor "turns current into a magnetic field", when current and magnetic fields are mutually related already.

It really should say that an inductor stores energy in a magnetic field, as a capacitor stores energy in an electric field.

 

[Throckmorton]

My friend is a developer at HP and for a while was a Comp Eng major, and he told me about this yesterday. He was excited about it. Since he knows what he's talking about, I think it's safe to say that this is big. It's not a freaking Segway

 

[uclaLabrat]

After the talk, my boss pretty much looked at me and said, "Yeah, that guy just won the nobel prize"

 

[Analog]

Originally posted by: George P Burdell
Sweet mother of Ohm, how will we perform RLC circuit analysis now?

Did anyone read the paper and understand how it is supposed to work in theory? Wikipedia says that at a given instant of time, it acts like a linear resistor; but the value of the resistor M changes with time depending on charge. So how exactly would one change M?

By the current going thru it. Did you look at the formula? Just like in a capacitor, as you introduce current through it, the voltage will rise until the current is removed, and then it "remembers" the voltage.

This device is similar. Its resistance changes as current is introduced through it. When the current is removed, its resistance stays the same, thus it "remembers" the resistance. Obviously, as you want to measure its resistance, your measuring device will deviate its resistance slightly, as you need a current to do so (so use a small current!). But this is exactly like a capacitor. To measure the voltage on a capacitor, you need to use a "voltmeter" or such, which has an intrinsic impedance that slightly degrades its voltage due to a small discharge (so uses a real high impedance!). In a DRAM, we need to refresh these "capacitors" on a regular basis, as reading the bits will discharge what's in them. This device will hopefully work in a similar fashion for memory storage.

 

[91TTZ]

Originally posted by: Throckmorton
My friend is a developer at HP and for a while was a Comp Eng major, and he told me about this yesterday. He was excited about it. Since he knows what he's talking about, I think it's safe to say that this is big. It's not a freaking Segway

Is your friend a veteran scientist at HP or just some generic developer straight out of college?

 

[techs]

Wiki has a article on it:

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

 

[manlymatt83]

Time to invest in HP?

Edit: yay, post 4444