Memristor, not memresistor.

Memristors are a fundamental component, analagous to resistors, capacitors, and inductors.

You probably wouldn't notice the difference, as computers work at a different abstraction level.

Memristors can function as both transistors and storage, so we could end up with a lot less concern about cache misses.

They can also mimic the function of neural synapses, so a neural net coprocessor is a possibility.

The scam of the Memristor. The Memristor debunked.

There are ppl in HP of all places, glossing about 'the incredible Memristor'.

According to them, this long elusive golden-fleece like 'component' will change the face of digital electronics and everything else.

I have bad news for them. The bad news is that the Memristor is overhyped, and over-their-heads. Memristor hype is resulting from a colossal mis-understanding due to the mix up of two ancient well understood phenomena. One: negative-resistance and two: analog-memory effects of passive components.

The problem is that the 2 effects are being mixed up in a really confusing bag, and being palmed off as a 'earth shattering invention'. This is truly sad.

Let me simplify the goobeldygook- the Memristor is supposed to be the 'complex-conjugate' of a the humble 0.1cent resistor. It is being advertised as 'the-opposite' in the sence that an inductor is the opposite of a capacitor. I dont know why no one is saying it in the following words "A Memristor is a negative resistance". Perhaps because it sounds too pedestrian to them, and they'd instead use hype.

My gripe with the intoxicated crowd is this- why is negative resistance newsworthy in any shape or form at all? We have had negative resistance in gas-discharge tubes [tube-lights and neon lamps] for over a 100 yrs. The pinched bow tie that the Memristor crowd is so exited about- its been right there all along in every tube-light ever installed. The Memristor crowd doesnt know that all gas discharges have negative resistance. We have been putting current limiting devices in discharge lamps for a long long time. Good old William Crook [Crooks Tube] and Tesla [Florescent-Tubelight] must be turning in their graves.

But wait, this is semi-conducting negative resistance,not messy old plasma, they say. Ever heard of the tunnel diode ? says I. Good old Esaki invented that one in the 60's. That's 50 yrs ago. Fifty ! Semiconducting Negative resistance is nothing new, and is not news worthy in any shape or form.

Not only gas tubes, and tunnel diodes, we also have Negisters. This is a 2 terminal passive device, that has been around for ages, which has all the negative resistance you'd ever want. Its basically a npn transistor, whose base is unused, and the emitter-collector are used in reverse. The emitter connected to Vcc and the collector to Gnd. The avelance breakdown gives negative resistance like behavior.

Then there is the Lambda Diode- a series parallel combo of 2 FETS, that gives negative Resistance. The phenomena is as old as the hills. But the Memristor crowd is not aware of it.

Stop, it has memory! they say. Daah, says I. Everything in nature has memory. All 0.1cent resistors have memory. Every single one. Ditto for 0.1cent disk caps. You just need to run them a little over their specified power rating. And they remember that abuse forever. It doesnt destroy the component and the change is easily measurable. Its hardly newsworthy at all. All analog designers have struggled with this problem forever. Make the resistances in bulk silicon, micron sized, and you have packaged analog memory.

But in true snake oil fashion, this spec-deviation 'memory-problem' is now being touted as 'a-feature'.

And in true debunker fashion, my gripe with that is - the memory effect has nothing whatsoever to do with the pinched bow tie, "golden-fleece-elusive-component-invention" part. Althought they are being packaged and presented together, the two are not related at all!

My suspicion is that the negative resistor that these ppl 'invented' has a side effect - memory ! And like most other analog world side effects, it will be hard to exploit favorably.

The problem is that the 2 effects are being mixed up in a really confusing bag, and being palmed off as a 'earth shattering invention'. This is truly sad. Really really sad.

The Memristor is not the revolutionary device it is being touted to be. But the hype machines are on. I deeply suspect they are looking for funding.

At best, this 'invention' might serve as a new kind of analog memory. Pretty limited in its applications and boring to boot!

ashvini new delhi india

I don't understand why your post implies that compilers produce... what? schematics of capacitors, resistors and inductors?