Slashdot Mirror
Nanocomputing Proof Point
untulis writes "HP Labs and UCLA researchers have apparently been able to produce logic gates via chemical processes rarher than standard lithography, making gates only a few atoms in size, according to Saturday's San Jose Mercury News. The article describes the gates as being a thousand times smaller than current gates. Mass production is at least a decade away, if the process turns out to be commerically feasible. "
"We experimented with laser scanning & optical techniques, traditional silicon metal layers, and other even more bizarre means -- in the end, we had to use silicon interconnect, which meant that the wires which connected to these devices were thousands of times larger than the devices itself."
Interesting that you mention that problem. I just received my August issue of Popular Science and they had an article about DNA conductors. A research team at University of Basel's Institute of Physics in Switzerland were able to ground one end of DNA and attach the other end to an electric contact. What they found is that the DNA can act similar to copper and conduct electricity. Not only that, they feel that there would be less scattering and slowing down of electrons with DNA "wire" than with traditional means.
And yes.. They hope that someday the DNA wire will be used in nanodevices for the connections.
Now that's scary. Having to run an SMP machine to be able to use Micro$oft's latest OS. I can see the minimum hardware requirements on the side of the box now. Four Pentium Pro 5 processors, 1 Gigabyte of shared RAM, and 2 Gigabytes of free hard disk space. (For optional voice interface add another Gigabyte of both memory and hard disk space.)
-- Some people say they can tell the time by looking at the Sun, but I have trouble seeing the numbers.
With Nanocomputers we are talking petabytes of available ram. You can squeeze a terabyte of ram into a structure the size of a sugar cube.
There are true limits to Moore's law, mainly the speed of light. Once we hit that brick wall its just a matter of throwing more processors and more innovative designs at the problem. We wont hit the end of computings advances for some time to come. Probably 10's of thousands of years.
competition-driven, you do NOT want to fully
publish your results until you have complete
proof-of-concept; especially in lucritive areas
like this, there are dozens of other labs (US
and int'l) watching for advances like this,
and they might be able to beat you to the
punch if you leak too many details. In this
case, this could cost the group a patent, a
grant, or other rewards for coming up with the
first proof-of-concept.
That's why articles like this, or the one about
teleportation being possible from about 6 months ago, or many other of the science articles being
posted to
group is strongly protecting their potental
assets.
"Pinky, you've left the lens cap of your mind on again." - P&TB
"I can see my house from here!" - ST:
Check out Nanosystems by Drexler and Merkle. It addresses these issues.
Actually, this is a real problem with nanoscale electronic devices. The answer possibly lies in mechanical nanocomputers. I've read scenarios that use chemical messengers between components, pressure waves, light, all kinds of things.
Interesting stuff nonetheless.
This is so kewl! I can't wait to get one of these babies so that I run Quake at 10000 FPS. And an overclocked dual processor system would be even more AWS0M!!!!!! True there aren't many details in the article, but I'm sure their irreluvant anyway. An if all those clewless researchers working on pushing silicon faster or on the hundreds of alternative new technologies wood just work on this one, we'd get it in only a few years!!!
With such raducl performance, all those complicated MP (massively polynomal) complete problems will be trivial! Imagine finding factors of a Hamiltonian circuit in just seconds!!! Course one of my teachers mumbled something about hyuge speed increases not really helping on MP problems, but that's obviously B0GUZ!!
The bad thing will be that this might keep Microsux with their bloatware in business, but that's OK cuz Linux will rock even more on this puppy. Forget those GNOME vs KDE flamewars; we'll have enough juice to run both at once!
David
Criticism of Goddard's Rocket Research: A New York Times editorial of 1921 said:
"That Professor Goddard with his 'chair' in Clark College and the countenancing of the Smithsonian Institution does not know the relation of action to reaction, and of the need to have something better than a vacuum against which to react -- to say that would be absurd. Of course he only seems to lack the knowledge ladled out daily in high schools..."
That may be true for UCLA (though I'd like to see the papers), but many universities participate in non-disclosable research (military, corporate, etc), in return for money/goods and the chance to attain data that they may not have been able to generate themselves.
;)
I wasn't saying that there was anything wrong with his behaviour, or that it is wrong for him to engage in research on behalf of an outside firm. The man is one of the most brilliant profs we have, and I'm glad he's stayed with the university rather than going out to make the millions I'm very sure he could.
Besides, my first guess is that the people at UCLA are simply completing their research so as not to make any premature announcements (ie. Cold fusion, etc.) Don't get your knickers in a knot.
Michael Gentili
- He's just some guy, you know?
Both of the other replies to this were quite insightful, and much appreciated. I have been out of the loop since 1995, after all. :) I'm glad to see people focusing on the interconnect issue, and wish that this news article perhaps took a bit closer of a look at the real technological challenges rather than those which have been solved in many creative ways prior to this research's publication.
"But always she's the spectre of uncertainty I first endured, then faded, then embraced..."
Friend of mine at UofT is working on that project. He's doing the grunt work; laying out DNA strands between conductors and running a current. Not having much luck so far, but here's hoping... :)
Michael Gentili
- He's just some guy, you know?
Silly me. here I was, thinking that because it was UNIVERSITY RESEARCH, the information would be, you know, PUBLISHED, say in leading scientific journals. Obviously I overlooked the fact that most scientific research at universities these days is WHORED OUT to the HIGHEST CORPORATE BIDDER.
Goodness, I'm so very glad we have people around to explain these esoteric concepts to the masses.
*spurt* indeed.
So what if it IS akin to building a computer with vacuum tubes? Did we not have to go through that stage with electronic computers? The technology will grow in stages, with new developments building upon older ones.
;-)
Perhaps we should have waited for the Pentium chip before building any PCs.
It was a joke! When you give me that look it was a joke.
If those organic molecules for logic gates can be built from common protein strand, the idea to genetically engineered E.Coli to make parralel computer is not far fetch.
...but will it run Linux?...*g*
We even worked on some transistors with I-V (that's current-voltage to you non EE types) curves which had two or three plateaus on them. Theoretically, this means that you could have a transistor which has three states instead of two (0, 1, and 2!)
The biggest problems in making these chips commercially viable -- oh, and by the way, we had Motorola fabricate the devices to our specifications -- were that most of them only exhibited their nifty behavior at low temperature (liquid nitrogen temps, or, if you were extremely unlucky, liquid helium temps of ~4 Kelvin) and that if you wanted to make an array of these things, you had to find a way of accessing all of them. Now, according to scientists like Prof. Reed, the temperature problem may be tractable through the use of high-temperature superconductive materials, but as for interconnect...
Traditional methods of accessing rows of transistors in memory cells don't work. You can't simply select a row and a column and expect the answer to trickle down to your buffers when the stored charge is a single electron. The same is probably true of a single molecule. We experimented with laser scanning & optical techniques, traditional silicon metal layers, and other even more bizarre means -- in the end, we had to use silicon interconnect, which meant that the wires which connected to these devices were thousands of times larger than the devices itself.
I hate to be a party-pooper, but until some sort of discovery occurs in the interconnect field, it won't matter if we can represent a binary state with a single lepton!
"But always she's the spectre of uncertainty I first endured, then faded, then embraced..."
I think It's time to train magazine and newspaper to also publish referance to people can hunt for further reading, just like in scientific paper.
I wouldn't get *too* excited about nanocomputing just yet.
Most announcements like this never make it into reality. Even if it does, it's still a very long way ahead. We've got many more exciting developments coming along much sooner, like cheap
SMP (the next big thing, IMHO). Even if we do get cheap, fast nanocomputers, Windows 2010 (or whatever) will still take just as long to reboot.
Who else thought of Dr. Evil's 'Mini-Me' when reading, "gates as being a thousand times smaller than current gates"?
Yellow tigers crouched in jungles in her dark eyes.
She's just dressing, goodbye windows, tired starlings.
Once the ball gets rolling on something like this, it tends to pick up speed beyond all expectations. Though I had projected 2015 before we started to see commonplace nano-apps...
Anyway, I expect we'll see real progress on this within 5 to 7 years. Might still be 2010 or 2015 before we start seeing consumers REALLY get into them. Anyone care to establish a betting pool for when we'll see the first diamond frame automobile?
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
I wouldn't get too excited about nanocomputing just yet.
Most announcements like this never make it into reality. Even if it does, it's still a very long way ahead. We've got many more exciting developments coming along much sooner, like cheap SMP (the next big thing, IMHO). Even if we do get cheap, fast nanocomputers, Windows 2010 (or whatever) will still take just as long to reboot.
Who else thought of Dr. Evil's 'Mini-Me' when reading, "gates as being a thousand times smaller than current gates"?
Yellow tigers crouched in jungles in her dark eyes.
She's just dressing, goodbye windows, tired starlings.
I don't know much about the inner workings of cpus, but anytime I someone suggests of a way to make Gates 1000 time smaller, it puts a smile on my head:)
"Uhmmm this might sound a little paranoid but, I want shielded twistedpair. I figure if I wear a tinfoil hat, my data s
The IEEE-CS of Cal Poly SLO was luck enough to hear Dr. Stan Williams of HP (one of the pioneers of "growing wires") give a talk on this very subject two years ago (if you get a chance to hear this man speak, do so). His description in post-worthy terms boils down to single electron gate operations where the electron is split in a wave form theory and either moves through the wire *switch on* or bounces back *switch off*. He continued to hint that they were producing raw waffers of these devices and etching them in a manner that can be described as "infecting them" with a virus and seeing what comes out on the other side because the devices on the waffers are simply too small to map. This waffer is functionally similar to a neural network, you don't quite know what's all there but you know how to use the cause and effect productivly.
Word has it as of this now two year old talk, that HP has already produced several functional 8 bit quantum processors that are capable of solving all results of a simple logic patern simulaneously.
By my own limited estimation, I'd have to belive that HP has a process in place to produce these waffers in high yeild but not at low cost yet.
Time tables on HPs commercial production are estimated between the year 2007 and 2012. Kiss your encryption schemes good bye
-panZ
--Let's hack root on 127.0.0.1 --panZ
This is really interesting stuff. If you wanted to do research on this what would most likely be your major in college? ChemE?
--Scott
I call him mini-Bill.
He'd still be a biter! Ha!
"Die Scott (McNeally)!" Haha!
- Ahem. Sorry. Shag flashback.
-- What you do today will cost you a day of your life.
I'm an idiot. The logic I posted is for an adder not a multiplier( which is far to complex for a simple explanation of what a logic gate does ).
Don't leave your mind so open that your brain falls out. Don't close it so much that you cut off the blood.
It sounds great, but I am sick of all these SLOOOW people, and semiconductor companies hacing to wait until current technology has been drained until bothering with the next generation of technology. I hope they know they wont be able to compete with some of the recent start ups, and I personally will be hapy seeing Intel and AMD, and Compaq squirming around for a new market as soon as optical FPGA array based computers are released, in LESS THEN A DECADE :P
Quantum computing is smaller than atoms.
The demo experiments use photons or electrons.
Quantum noise can mess up electronics of a just
a few atoms thick, however.
Its a lot like running out of oil - always 10-20
years in the future. People were whining in the 70s that Moore's law would end by Y2K. Technology seems to keep the limit out there.
It's not just UCLA, HP labs is involved as well. They are possibly running with HP funding and therefore under HP Non-Disclosure. I do research at UofT, and while Canadian Universities are less dependent upon corporate grants than American Universities (That's changing, unfortunatly), there is still a strong relationship between any lab and it's corporate sponsor.
.2 microns wide, and the N-doped wells about, oh .4 microns wide, etc." So someone in the class, frustrated with these vague numbers asked for the exact numbers. And the guy gives her this pained look and says, "Uh, well, I'm actually doing some development right now with a die fab (chip manufacturing firm) and I can't release the numbers, I can't even tell you what firm I'm working for."
;)
In fact, during a VLSI course this year, a professor at UofT was speaking about transistor sizes in chips and was giving us rough sizes like "Poly lines can be, oh around
Pretty spooky for an "open academic institution" eh?
Also, it is just possible that they're not sure of their numbers yet. The scientific community is merciless with those who release numbers that aren't rock solid. They're probably just covering their asses. I would.
Michael Gentili
- He's just some guy, you know?
We don't know anything about the tech. There's no reason for the gates to run hot. A superconducting pentium II would run at just above room temperature, if it existed :). So before we know any more about the substances and their characteristics, we can't say anything about them.
Michael Gentili
- He's just some guy, you know?
Arggh!
Repeat after me: You can't send information this way. You can't send information this way. You can't send information this way.
Imagine the following scenario:
- You have a bag containing a black marble and a white marble.
- Two people pick marbles with their eyes closed and then,
- Without looking walk a far distance away and then look at their marble
Instantly, they'll know what color marble they have and the other person has. So what. Not one single bit of information has been transmitted from one person to the other; all that's happened is that equivalent bits of information have been sent from the bag to each person. And since it was a random bit of information to begin with, you're no better off than before."...you could have a transistor which has three states instead of two (0, 1, and 2!)"
yeah, sure. that would work REAL well with the logic computers are based on. programmers would have to deal with booleans with values like true (1), false (0), and whatever (2). Think of how the numbering systems of computers would change... there could be 6561 (3^8) in 8 bits instead of 256 (2^8). hey... then we wouldn't need two whole bytes for all those chineese characters! It would take a while to change the 1,000,000,000 decibytes (or however much it is) of data there is stored on computers from binary to trinary, and processors would have a hard time with the logical operaters (and, or, xor, not, etc.) What's 2102012^12012001 in base 3? computers would have to convert from trinary, to binary, then to trinary (to be transferred) and then finally to decimal, instead of binary to decimal.
Why is it like that? Chips are created NOW with a chemical process. It's just a different one. What do you think lithography is?
If you're referring to us retracing our steps to branch into a new field, well, yes. Every application of a technology has its functional limits. And I hope we're well on our way to developing a new computer tech before we reach the limits of the old one.
Michael Gentili
- He's just some guy, you know?
Ah, but consider the applications for generating random encryption keys.
If that's all there was it wouldn't be very interesting.
However, some research summaries I've read (sorry that I can't give references) imply that there is a way to see (measure) a photon in the polarization of your choice.
To extend your analogy, this would be where person 1 chooses to see his marble as white, which would then force person 2's marble to be black. So person 1 has just transmitted a bit of information to person 2 instantaniously.
AFAIK, no one's proven that this can actually happen, so it may be in the same category as travelling back in time. It's just some interesting speculation.
Nanotech is the cybernetic transformation of molecular manufacturing.
What does this mean?
simply "grow" things from the bottem up.
like a plant can grow out of a seed, by dissembliing and assembling molecules.
The natural universal assembler is called a ribosome. It builds all proteïnes in a biological substance given information of DNA.
So what I mean is: do trees get hot?
PS: The Future=Software
The applications of this in miniturization of microchip technnology should be pretty interesting. Imagine thumbnail size motherboard
Oh. First post?
Don't leave your mind so open that your brain falls out. Don't close it so much that you cut off the blood.
I think we probably need to hear more.
Having proved that the concept might work...
Heh.
This sig is false.
thats about all i can say, WOW
I don't know a whole lot about computers, so umm,
what exactly do the gates do? is it the little
'yes' or 'no' switch type thing for each bit of
memory?
More smoke in the water.. do journalists just hang around universities waiting for people to make "breakthrough's" so they can go and not publish the results. How much stuff like this gets published into scientific journals, with way more detail than this, and never makes it onto places like slashdot? Surely people working at UCLA should be a little more open about their research than, say, Xerox/PARC. That assumption appears to be false.. or is the popular media just jumping the gun a little.
Oh and BTW.. if it aint molecular manufacturing (that's little nanobes that can self-replicate and make a wide range of ridged structures.. see nano.xerox.com/nano/) then it aint a breakthrough in nanotechnology.. mildstone, maybe.
How we know is more important than what we know.
Posted by Napalm4u:
Man I hate this!! I won't see this for 10 years! man that's crazy. Get some government grants and make them, I want to overclock it now!
btw ll:22 posting! my clock is off 20 minutes!
They don't say anything about the speed at which these can be manufactured. That's been the biggest problem with many of the non-lithographic approaches mentioned so far--it simply takes too long for most processes to produce a full chip with a reasonable amount of equipment. Anyone know any more about this technique?
Now all they have to do is train some of those otherwise worthless E-Coli strains to exhude these gates from their butts.
Then, what the hey, send some of these E-Coli to oh, Jupiter and let them turn the whole planet into a single CPU. Now that would be the shits.
Look closely - they are possibly using DNA to bind the chemicals in the right places... another great use for what nature has given us... and great potential for a few things, like fingerprinting each chip, allwing for built in neural circuits, and easy integration into one of us...
wow, i say...
On one hand, a processor like hat would possibly double as a space heater (or soldering iron... :-)
think about it: let your computer ray trace an image as you hold it to build a ham radio
On the other hand, giving up speed for parallelism
would make mahines with this tech cold and dreamy.
The lack of details is a bit disappointing - I think this achievement is more a proof of concept of manufacture, than a validation of nanotechnology as a whole. Also, building a nanocomputer with a chemical process seems akin to building a computer with vacuum tubes...
well, this thing is so new, noone has a category yet.
Duke has a cross disiplinary grad program for nanotech. It's mostly chem/material science/molecular bio.
but great majority of people who so nanotech are chemist/solid state physic people.
nanotech is like biochem in the late 60's. It is embryonic and way cutting edge.
So you grow the processors, build the memory is steel, mix it all up and you have a brain. With deterioration, incoherent thought, and possible some pretty fierce AI. 10 years, eh? hmmmm....
+&x
I read an article a few years ago about a "computer" built from molecules of chloroform. I think that it was tested by searching through a snapshot of the web to find a particular word. Bookmarks come and go, but I found a few articles on the subject:
A Vision of Synthetic Prophecy
Quantum Computing with Molecules
The computing aspect of this is really cool, as it would make factoring Really Large Numbers a snap (because these computers would be massively parallel and would be execute many instructions in one step). These machines would have the ability to factor a 400 digit number in about a year. The networking applications of quantum computing are pretty interesting as well. If you can create two photons with opposite polarization, as soon as you measure the polarization of one, the state of the other is immediately fixed, regardless of the distance.
I go to school with an engineering scientist who did his thesis on quantum computing. What these people are looking at doing, is producing a new method of doing the same old thing, building ever smaller and smaller transistors to cobble together into a CPU.
;)
Quantum computing chucks the transistor out the window and works directly with molecules and electrons and probability and stuff. I'm just a computer engineer, I understand transistors and circuits and things with stuff and the resistance. All this new fangled probability computing scares me
Actually, it's really cool, nano tech will make it much more practical to produce quantum computers, so we can abandon the transistor model, and get on with some serious parallel computing!
Michael Gentili
- He's just some guy, you know?
HP Labs and UCLA researchers have apparently been able to produce logic Gates via chemical processes rather than standard lithography, making Gates only a few atoms in size, according to Saturday's San Jose Mercury News. The article describes the Gates as being a thousand times smaller than current Gates. Mass production is at least a decade away, if the process turns out to be commerically feasible.
;-)
:-)
To sum it up: If this works out well, Linux will soon be overrun by lots of little Bills, giving Microsoft a totally new meaning...
Sorry - couldn't resist!
-- Eavy (: Linux Is Not UniX
The "writer" of the article hinted that without advances such as these, chip makers would be able to extend Moore's law "only for another 10 to 15 years at most."
Not that I am in any way looking forward to technological stagnation, but wouldn't it be interesting if we did hit a brick wall for a few years. All those surplus design elements (dancing paperclip, configuration wizrds, the gui itself) would suddenly be seen as the drain they actually are.
Software doesn't run fast enough? Upgrade to the latest model. But what if the latest model chip was just as fast as the one you have now?
I know, I know, more RAM, multiple processors, Beowulf clusters, but it would still be interesting to see the effects of a problem like that.
Judging from the content of that post, it should be no surprise that it was moderated down as it did not appear the poster was willing to contribute significant content. Quit yer whinin!