HP responds to memristor debate

1/20/2012 9:20 PM EST

Very interesting indeed. So, it sounds like the pinched hysteresis curve is the only requirement for this memristor, and anyone who thinks that there MUST also be a flux and charge relationship is simply being too hard-core literalist about it? Could be.

I can dream up a memristor without too much trouble. Just imagine sending a large AC current through a regular resistor, more than the resistor is "rated" for. The resitor heats up, changing its resitance value as current becomes greater.

Then the sine wave goes past its peak, and the current value decreases.

As the resistor cools, its resistance again changes, but not at the same rate as when the resitor was heating up. Assume that when the current was going up towards the peak value, and reached a value of .9i, the resistance was R1. When current started descending from the peak, and again reached a value .9i, the resistance of the now-cooling resistor is R2.

Not so hard to imagine. For example, it's easy enough to think of the resistor being slightly cooler when the current was increasing, and slightly hotter, at a given current value, when the current is decreasing.

The hysteresis curve is pynched, because this is a resistor, so when v=0, i=0, simply because there is no stored charge or magnetic flux.

And the next also-not-too-difficult conceptual exercise was to imagine similar properties for capacitors and inductors. Values of C and L that vary over time, as voltage is applied.

1/21/2012 7:53 PM EST

The curve doesn't tell you how it happened, there are volatile and nonvolatile means.

1/21/2012 9:08 PM EST

One has to chuckle at the PR machinery.
That is not a response at all, merely a (poor) deflection.

1/24/2012 8:01 AM EST

Ever since the tunnel diode revolution, I don't get too excited about new technology until I can buy it downtown.

1/24/2012 4:48 PM EST

I do not believe that was intended to be a deflection, just something along the lines of "We would agree with you if you were right" .... or the academic equivalent of just because you don't understand something does not mean that it is not true.

Question though, is there a difference between a dynamic system (complex) and a single element that display a certain characteristic. I can design a complex circuit element that looks to the external world like a resistor. That does not of course make it a resistor. Debate?

2/15/2012 10:18 PM EST

My rebuttal to HP's rebuttal is available at

http://www.scribd.com/doc/79648334/Memristor-Scientific-Method