I've been reading up on this crossbar latch that HP has been working
on. I think that while individual Memristors need to come out as
discrete components the same way as resistors, capacitors, etc do, we
shouldn't discount the crossbar latch itself. They speak of it
working in both analog and digital modes. A grid of memristors in
such a device could have amazingly useful potential for more advanced
BEAM applications. I envision linking sensors to control lines on
these arrays, and creating wide networks of Memristor "synapses".
This may be exactly the device that BEAM has needed but not had to
I can't wait until both individual Memristor components and crossbar
latch chips are released.
Begin forwarded message:
> From: "antti10001" <antti10001@...>
> Date: August 27, 2008 5:17:36 PM CDT
> To: email@example.com
> Subject: [beam] Re: MEMRISTOR , New electronic circuit Element,
> Reply-To: firstname.lastname@example.org
> --- In email@example.com, Richard Caudle <frankendaddy@...> wrote:
> > So the question arises...how do we make them and where do we find
> materials? :P
> Well, the Memristor made by HP Labs don't use platinum iridium alloy,
> but apparently titanium dioxide, which is much more common material.
> (E.g. in toothpaste).
> See the Wikipedia-article:
> and then read EE times and IEEE Spectrum articles
> linked at the end of page (they are both good).
> I just wonder when memristors are available
> as macroscopic components (i.e. not as a part
> of some gigabyte-sized exotic memory chip)
> if ever.
> One particularly nice feature is that they "remember
> their value", i.e. are not volatile. That would be useful
> in a solar-powered robot, that tries to accomplish some
> "long-term" task requiring multiple cycles of wake-ups and
> sleep. Also, analogish, not just digital.
> Now everybody, start writing patent applications where some
> simple (or complex) task has been implemented with the help of a
> memristor. Or publish your ideas here instead, as a "prior art",
> foiling the plans of evil patent lawers...
> > --- On Tue, 8/19/08, wrigter <wrigter@...> wrote:
> > From: wrigter <wrigter@...>
> > Subject: Re: [beam] MEMRISTOR , New electronic circuit Element,
> > To: firstname.lastname@example.org
> > Date: Tuesday, August 19, 2008, 10:26 AM
> > This is not some new IC but a scientific breakthrough. It is
> > the discovery of a new fundamental electronic circuit
> > element whose existence was predicted 37 years ago. The
> > Memristor is the stuff of Asimov's fabled Positronic brain
> > that used a Platinum Iridium alloy as a substrate for neural
> > networks. Very beamish!
> > wilf
> > ----- Original Message -----
> > From: Gary T <grogyan@woosh. co.nz>
> > To: beam@yahoogroups. com
> > Subject: Re: [beam] MEMRISTOR , New electronic circuit
> > Element,
> > Date: Tue, 19 Aug 2008 18:26:37 +1200
> > >Those things have been out for years.
> > >
> > >They are called digital pots, still too big for me in
> > >miniature robotics The only way that I can use them is if
> > >they are in 0603 or 0805 low profile type packages.
> > >
> > >Gary T
> > >
> > >>Hey all, i might be a bit late on this, in terms of dates
> > >>since around April, but a new electronic circuit element,
> > >>the Memristor, has been created.
> > >>The most interesting characteristic of the device is that
> > >>it remembers the amount of charge which flows through it.
> > >>
> > >>The resistance of a memristor depends on how much charge
> > >>has gone though it. If charge flows in one direction the
> > >>resistance will increase. However, if the charge now flows
> > >>in the opposite direction it will decrease.
> > >>
> > >>These things can probably be used as digital switches or
> > >>new analog devices. Perhaps used in Nv memory and neuronal
> > >>computing?
> > >>
> > >
> > >
> > >