Tech Talk iconTech Talk

Memristors: Coming soon to a brain near you.

HP Labs has wrestled the memristor into submission. Or, to put it less hyperbolically, by engineering control over the memristor, they've made it possible to design integrated circuits that incorporate the devices.

EETimes reported earlier this week that memristors, the fourth element of circuit design theory, will likely be the basis for HP's 2009 RRAM (resistive random-access memory) prototype chips. RRAM, a new kind of non-volatile memory that could be faster than flash, is also being pursued at other goliaths like Sharp, Samsung, and Fujitsu. But HP's plans go way beyond low-power chip design. Analog memristors, the researchers speculate, are a promising route to chips that operate more like the human brain than like a computer.

"RRAMs are our near term goal, but our second target for memristors, in the long term, is to transform computing by building adaptive control circuits that learn," said [HP Labs principal investigator for memristors Duncan] Stewart. "Analog circuits using electronic synapses will require at least five more years of research."


The same picture you will see in every article about memristors for the foreseeable future.

Back in April, HP anounced their successful manifestation of the hypothetical fourth passive circuit element, the memristor, first proposed by Leon Chua in a 1971 paper. The memristor joined the existing holy trinity of circuit elements-- resistors, capacitors and inductors-- with implications that were frankly kind of staggering. At the time, Stan Williams told me that they already had a chip humming away in the lab-- this wasn't just a bunch of hypothetical design specs. But the discovery met with much skepticism:

There is nothing new or revolutionary here. Bulk resistance versus incremental resistance is well known. Nonlinear resistances are well known. Resistance which varies depending upon the history of the device are well known. The claim that a fundamentally new circuit element has been found is overblown.

Similar fightin' words could be found in an uncharacteristic slapfight that broke out in the Spectrum Online comments section.

But now, EETimes reports (in what is by the way a must-read article) that HP labs "has demonstrated how to control its memristor material, which changes resistance in response to current flowing through it:

"We now have experimental proof that our memristor behaves the way our theory predicts," said [Stewart]. "Plus, we have now demonstrated engineering control over the memristor device structures, which will enable us to build real chips very soon."

HP Labs maintains that because of this property, massive memristor arrays could enable brain-like learning.

In the brain, a synapse is strengthened whenever current flows through it, similar to the way resistance is lowered by flowing current through a memristor. Such neural networks could learn to adapt by allowing current to flow in either direction as needed.

In case I haven't infuriated enough people, I will now half-jokingly mention the Singularity.

US Air Force Tanker Bid to Be Reopened

The US Department of Defense reopened the bidding between Boeing and the Northrop-Grumman/EADS/Airbus consortium for the midair refueling tanker contract yesterday afternoon because of major flaws the Government Accountability Office (GAO) found in the original source selection. Secretary of Defense Robert M. Gates made the announcement, and took away the source selection authority from the US Air Force and gave it instead to Mr. John Young Jr., undersecretary of defense for acquisition, technology and logistics.

Acting Secretary of the Air Force Michael B. Donley said that even though the â''GAO's conclusions show that even in a large, complex procurement with considerable staff resources and oversight, work accomplished by our contracting personnel, our warfighters and our engineers is not always adequately prepared to withstand the detailed audits and the legal challenges that we can now expect,â'' Air Force acquisition is not â''fatally flawed.â''

The Air Force will still get to manage the program after the contract is awarded (and hopefully not protested).

In a bit of jaw-boning (or wishful thinking), Young said, â''I think that would probably be the only silver lining in this, is the possibility that both teams decide to sharpen their pencils and offer the taxpayer and the warfighter an even better deal.â''

Young better be careful for what he wishes for. â''Optimistic biddingâ'' â'' or buying-in â'' is a major reason for procurement messes, as the GAO once more recently pointed out.

Out of Africa: Cooking with Human Waste

The shanty towns of Africa's great cities are incubators for material innovations that easily escape the ambitions of people living in wealthier surroundings. In Nairobi's Kibera slum, the latest sensation is a "bio-latrine" that converts human waste into energy that produces gas for lighting and domestic cooking.

The technology behind the bio-latrine was developed by the governments of Kenya and France. Under a project supported by the United Nation's Habitat agency, about 20 bio-latrines that convert waste into gas will be installed in four areas of Kibera.

The bio-latrine consists of a shallow pit latrine, a "bio-digester" that produces gas and fertilizer; and a dispenser. Above the latrine are toilets, a kitchen and a community meeting room.

The bio-latrine requires cooperation among neighbors -- and that may be its undoing. Crime in Kibera runs high; so does envy and conflict between neighbors. Yet the problem of human waste in African shanty towns is large and growing. Informal settlements in cities can't wait for government to act. Small-scale technologies, under the control of ordinary people, provide one answer to the glaring need in Nairobi -- and other African cities -- for better sanitation and less costly fuel for cooking.

While the bio-latrine is surprisingly complicated for an "appropriate" technology, it does address the two most vexing problems of poor Africans in makeshift neighborhoods and even schools. And that's reason enough for more experiments with them.

A Website Devoted to Nanotech Conferences...the Apocalypse is Nigh

For most involved in nanotechnology, even tangentially, one must learn early on the survival technique of how to sort through the truckload of conferences that are held weekly throughout the world, or even how to avoid them entirely.

The easiest way has been to use various websites like Nanotech-Now that provide a pretty thorough list of everything going on. There are many other similar sites that I will not mention simply because I am sure I will forget one of them.

But now, for those who are somehow challenged by using any number of these other conference listing sites, there is a new website solely dedicated to listing conferences.

What I enjoyed in the press release announcing this website was this tidbit â''This is believed to be the only website exclusively devoted to meetings in the rapidly emerging field of nanotechnologyâ''. They could be wrong, but this is what they believed at the time of the press release. Nicely done. I am sure the lawyers were pleased with that little tap dance.

When the two great hyped â''industriesâ'' of the last decade align in such a senseless way, we are hard pressed not to see this as a sign of dark days to come.

Plug-in Priuses: Toyota Dealers Getting Cold Feet?

Just a month ago, we told you that Hymotion named 4 Toyota dealers as official installers for its $9,995 plug-in conversion module for 2004-2008 Priuses.

Now, in a comprehensive piece in Green Car Advisor, the blog on all matters automotive and green, Scott Doggett writes that two of the four seem to be getting cold feet.

There's a continuing story with much left to be played out here, but we wanted to bring you the latest update as soon as we could.

And what, dear readers, are your thoughts? Should Toyota permit its dealers to install Hymotion's kits--assuming the company can back up their claim that the kit passed all relevant new-car crash tests? If you chose to install a plug-in kit in your Prius, how much would the liability issues worry you?

A Little Bit of Knowledge is a Dangerous Thing When Discussing Nanotech

I came across this tidbit that gave me a bit of chuckle.

Apparently, itâ''s an excerpt from a book entitled â''The Design of Life.â'' I am guessing that the book is supposed to be a kind of defense of â''Intelligent Designâ''. But I am not sure. I canâ''t be bothered with that kind of drivel to actually find out.

But the reasoning process used in this little snippet is hysterical. Apparently, skeptics believed nanotechnology would never work because â''such small machines are unworkable.â'' Umh, should someone put him in touch with the Foresight Institute and the National Nanotechnology Initiative just to get his definition of nanotech straight? Nahâ'¿

It turns out the skeptics reached their close-minded perspective based on â''Heisenberg's Uncertainty Principleâ''. Yikes! Gee and I always thought it was based on Van der Waal forces and Brownian motion, nothing quite as esoteric as quantum physics.

I am beginning to think there may be a new law in physics that states that anyone who uses the term â''Heisenbergâ''s Uncertainty Principleâ'' and is not more than well acquainted with physics is talking a lot of nonsense.

I understand that this is the norm in public debate, but for me itâ''s just a pity when nanotech gets dragged into it.

Overcoming the Funding Gap in Nanotech

About five years ago I helped organize a conference that contained a panel session called â''Bridging the Funding Gapâ''. At the time, this was a matter of some concern. How were innovations in nanotech going to move from government funding at research institutes and universities towards commercial markets?

The answers were not clear at the time, and now while things may be a little clearer they may be less hopeful.

This state of affairs is highlighted in a recent post over at TNTLog â''Can the VC Model Handle Emerging Technologies?â''. The answer seems to be an unequivocal â''noâ''.

As pointed out in the piece, the Venture Capital model (a 40-year-old model as pointed out in the article) requires some kind of financial exit within seven years of the initial investment. With early-stage companies developing emerging technologies, and in particular nanotech, seven years is just about where things get anywhere near markets.

The landscape today for many of these early-stage nanotech companies is like coming across one carcass of bones after another in the desert. Some of the brightest stars of nanotech five or six years ago have had their IP portfolios sold off for as little as $1,000.

To some extent, the funding gap question has been answered in that large companies are now spending the most money in R&D and innovation in nanotech. In other words, we are more likely to see the cure for cancer, the next breakthrough innovation in electronics, or a viable alternative energy source come from the labs of large industrial company than from a spin-out from a university lab.

But this model certainly does not cover all the possibilities for innovation, and will likely preclude important breakthroughs made outside of it to make it to market.

But the financing model for this type of innovation clearly does not seem to be the VC model. Whatever model is finally used, it is clear that it will need to stay in that investment for a lot longer than seven years.

US Air Force Tanker Acquisition Fiasco Is Difficult to Explain

The Government Accountability Office (GAO) yesterday released its detailed (although redacted) reasons as to why they sustained Boeingâ''s protest of the award of the new USAF refueling tanker program to the Northrop Grumman-led team. It makes for grim reading.

The GAO found that â''(1) the Air Force did not evaluate the offerorsâ'' technical proposals under the key system requirements subfactor of the mission capability factor in accordance with the weighting established in the RFPâ''s evaluation criteria; (2) a key technical discriminator relied upon in the selection decision in favor of Northrop Grumman relating to the aerial refueling area of the key system requirements subfactor, was contrary to the RFP; (3) the Air Force did not reasonably evaluate the capability of Northrop Grummanâ''s proposed aircraft to refuel all current Air Force fixed-wing, tanker-compatible aircraft using current Air Force procedures, as required by the RFP; (4) the Air Force conducted misleading and unequal discussions with Boeing with respect to whether it had satisfied an RFP objective under the operational utility area of the key system requirements subfactor; (5) Northrop Grummanâ''s proposal took exception to a material solicitation requirement related to the product support subfactor; (6) the Air Force did not reasonably evaluate military construction (MILCON) costs associated with the offerorsâ'' proposed aircraft consistent with the RFP; and (7) the Air Force unreasonably evaluated Boeingâ''s estimated non-recurring engineering costs associated with its proposed system development and demonstration (SDD)."

These are non-trivial errors, and not easily understandable given the highly political and extremely controversial nature of the acquisition (see here for a good summary of the history of this procurement), or that the US Air Force senior acquisition officials claimed that this was its most thorough, transparent and fair acquisition ever.

To wit, at a news conference announcing the award on 29 February 2008, then Secretary of the Air Force Michael Wynne said, â''Today's announcement is the culmination of years of tireless work and attention to detail by our acquisition professionals and our source selection team, who have been committed to maintaining integrity, providing transparency and promoting a fair competition for this critical aircraft program. They took the time to gain a thorough understanding of each proposal. They provided continuous feedback on the strengths and weaknesses of each proposal, and they gave the offerors insight into the Air Force's evaluation.â''

Mrs. Sue Payton, the Assistant Secretary of the Air Force for Acquisition, added at the same news conference that, â''â'¿ I can't stress enough what an incredibly open and transparent and rigorous first selection we have gone through. For months and months, we have been telling each offeror where their weaknesses were, where their strengths were, and so they've had a lot of opportunity to communicate with us and to make sure we were not talking past each other.â''

Mrs. Payton also added that, â''There was absolutely no bias in this award.â''

The GAO report seems to directly contradict many of these statements. I cannot fathom how the Air Force so badly messed up this acquisition. The Air Force knew that it was going to be scrutinized with a fine tooth comb, so everything they did had to be as Caesar's wife, above suspicion.

The questions being raised - fairly or unfairly - fall into "there was overt bias involved," which would be devastating for the Air Force's credibility and its senior leadership, or as the Brits say, it was a major cock-up. Neither explanation reflects well on the Air Force.

The Air Force has 60 days to respond to the GAO recommendations, which are that the â''Air Force reopen discussions with the offerors, obtain revised proposals, re-evaluate the revised proposals, and make a new source selection decision, consistent with this decision. If the Air Force believes that the RFP, as reasonably interpreted, does not adequately state its needs, the agency should amend the solicitation prior to conducting further discussions with the offerors. If Boeingâ''s proposal is selected for award, the Air Force should terminate the contract awarded to Northrop Grumman."

Congress really needs to dig into what happened here, and determine whether Air Force acquisition, which has been teetering on the edge of disaster for over fifteen years, is now totally broken.

Worst of all, of course, is that Air Force pilots are increasingly be placed at risk because the Air Forceâ''s aging tanker fleet â'' the average age of the KC-135 tankers is over 46 years â'' is going to take even longer to replace. The replacement process started in 2001, and here it is mid-2008. The lack of tankers is hindering Air Force operations, even though the Air Force likes to pretend the impacts are minor. My friends in the service are telling me otherwise.

I have been working a story for over a year on the problems in defense acquisition that is scheduled to appear later this autumn. It is not an altogether happy one either.

Airlines are running out of gas. Literally.

airstatraversecity_fueltruck_75.jpgHigh fuel prices are hammering the airline industry. I know that, and that all these annoying surcharges recently announcedâ''for luggage, for drinksâ''are attempts to somehow make up for the extra fuel costs.

What I didnâ''t know until yesterday was that airlines are also saving money on fuel by simply not filling up the tanks as high as they ought to. The FAA requires that an aircraft carry enough fuel to reach its destination and its most distant alternate airport, plus 45 minutes worth. Itâ''s not to the airlineâ''s advantage to carry any more than the minimum requirementâ''more fuel means a heavier plane; a heavier plane gets worse mileage.

Of course, the calculations are based on estimatesâ''estimated passenger weights, estimated luggage, estimated speed. And since my experience yesterday, Iâ''m thinking the airlines are estimating a little low.

I flew from San Francisco to Newark on a Continental 737. Great weather on both ends, only a little turbulence. A lovely flight, actually, best Iâ''ve had in a long time; the flight attendants were cheerful, passengers got two drink services and a bagel-and-egg sandwich without charge.

A little more than four hours into the flight, the pilot reported that weâ''d be landing about half an hour ahead of schedule, and flight attendants began collecting trash in preparation for our approach. Perhaps ten minutes later the pilot announced that weâ''d have to slow down a little to get in line for landing, but weâ''d still get in well ahead of schedule. I was thrilled; this would be my first flight in at least a year that landed on time, perhaps I could call a friend for dinner.

And then, just a few minutes later, the pilot came on the P.A. system again. â''Uh, folks, weâ''re going to make a quick stop for refueling.â'' Huh? Passengers looked at each other in surprise. Flight attendants passed rapidly through the cabin checking seat backs and tray tables and strapped themselves in. Minutes later, we landed at Stewart Air National Guard Base, less than 100 miles from our destination. After a long taxi past National Guard cargo planes, we parked and waited for the fuel trucks.

I was flabbergasted. Iâ''ve logged a lot of airline miles over the years; Iâ''ve never been on a flight that ran out of gas. The pilot blamed the problem on air traffic delays; but from what he had said the delay seemed minimal; certainly not enough to eat up the

45-minute reserve the FAA requires. I wondered about the fact that weâ''d gotten so far ahead of scheduleâ''likely weâ''d been flying a little faster than is optimal for fuel conservation. I also looked around me at the planeload of European tourists heading home, and wondered if theyâ''d been taking advantage of the cheap dollar and ended up bringing back a lot of extra luggage.

Turns out, though, that the Department of Transportation recently singled Continental out for having the most of what it calls â''minimum fuel declarationsâ'' into Newark airport last yearâ''96, more than twice the amount it made the previous year. Declaring â''minimum fuelâ'' flags the air traffic controllers that, if incoming planes need to be delayed, they shouldnâ''t delay this particular flight too much. The Department of Transportation also found out that Continental has been pressuring pilots to cut back on the amount of fuel they carry, for ultimately the pilot makes that decision. In October, the airline sent out a memo to pilots pointing out that â''adding fuel indiscriminately without critical thinking ultimately reduces profit sharing and possibly pension funding.â'' Gee, ya think thatâ''s why I ended up sitting on a deserted runway for an hour and a half?

Next time, Continental, forget the bagel, just fill up the tank.

Imperfect is just Perfect for Nanotubes in Solar Cells

Dye-sensitized solar cells, also known as Grÿtzel cells for one of the their co-inventors Michael Grÿtzel of the Ã'cole Polytechnique Fédérale de Lausanne, have been a promising alternative to silicon cells since 1991.

Basically, they use an organic dye that captures the incoming photons from sunlight to produce excited electrons.

The design of these cells typically includes a transparent, conductive coating that serves as the anode. The coatings are usually made from an oxide such as titanium dioxide and are placed on glass. Using glass has its obvious limitations and alternatives such as platinum are prohibitively expensive.

As a result, researchers have been experimenting with carbon nanotubes to replace these films. It was necessary that the CNTs achieve optimal transparency, conductivity and catalytic function. Unfortunately, instead of being optimal, they were more in the mid range for these areas.

But researchers have found that if they the give the CNTs imperfections by exposing them to ozone, their catalytic function improves dramatically.

More efficient and cheaper are the demands of renewable energy alternatives. Nanotechnology research seems to be listening.


Tech Talk

IEEE Spectrum’s general technology blog, featuring news, analysis, and opinions about engineering, consumer electronics, and technology and society, from the editorial staff and freelance contributors.

Newsletter Sign Up

Sign up for the Tech Alert newsletter and receive ground-breaking technology and science news from IEEE Spectrum every Thursday.

Load More