Slashdot Mirror
Scientists Build Computer Using Carbon Nanotubes
trendspotter writes "Future computers could run on lab-grown circuits that are thousands of times thinner than a human hair and operate on a fraction of the energy required to power today's silicon-based computer chips, extending 'Moore's Law' for years to come. Stanford engineers' very basic computer device using carbon nanotube technology validates carbon nanotubes as potential successors to today's silicon semiconductors. The achievement is reported today in an article on the cover of Nature magazine written by Max Shulaker and other doctoral students in electrical engineering. The research was led by Stanford professors Subhasish Mitra and H.S. Philip Wong."
Hasn't this been done before?
Current computers could run on lab-grown circuits that are thousands of times thinner than a human hair
Making a claim like "one day a computer will be thinner than a human hair!!! OMG it'll be great!!!" will just make you sound like an idiot sooner than you think. Lots of the quotes about computers fitting in single rooms and doing thousands of calculations are just like this.
Doesn't take much to excite you, doesn't it?
The most interesting thing about these alternative transistors might be environmental impact. I'm under the impression that traditional wafer fab is water intensive and heats and/or pollutes water. There are dangerous things such as arsenic and bromine involved. If the carbon nano-tube process is clean that'd be awesome. It would be great to think that we could dispose of obsolete technology by incinerating it, and not release anything other than CO2 into the air, leaving behind slag that's full of recyclable silver and copper.
For all intensive purposes, "whom" is no longer a word. That begs the question, "who cares"?
I knew my old tube set would be back in style again!
Time Bomber the Book coming soon.
I can't remember the book I read this in, but it posited that if you remove the silicon part of Moore's Law and you just talk about computing power and cost and the like that you can make a case that it has been in place throughout human history. In other words computing power has always been doubling, it just started by drawing numbers in the dirt, went to the abacus, etc.. etc... until we reached the silicon age and integrated circuits.
The hand wringing that the idea behind Moore's Law will ever end is just silly. When we reach the limits of silicon chips some other technology will take its place. This is just how human technology works.
If only we found a way to manufacture them.
1. "Lab-grown circuits that are thousands of times thinner than a human hair" is exactly what one could use to describe current silicon circuits. In fact, this study made transistors that are a micron across (which is, at best, hundreds of times thinner than a human hair), compared to current state-of-the-art silicon which is in the 22-28 nm range.
2. "A fraction of the energy required" does not describe the current study, nor was it their intent, from what I understand about the researchers' claims.
That's not to say that the research isn't very valuable; it looks like the level of integration they've managed is significantly better than what anybody else has achieved. But at the same time, there are lots of other ways that you could build a circuit that uses more area, costs more, takes longer to build, and is less power-efficient - this is just one more. All they've demonstrated is that you can hook together more than a handful of transistors successfully - but nowhere near the billions that they'd need for a commercial product.
The real breakthroughs have yet to be made; making it cheaper, smaller, faster, more efficient, and easily manufacturable - all at the same time. Not until all those problems are solved will it even have a chance of replacing real silicon. Until then, this is yet another case of a university PR rep boasting about their institution's research with grand claims about what the future holds, while not really reflecting the true nature of the research at hand.
(Admittedly, it is more boring when you adhere to reality.)
If nanoscale tech is available to universities now, what have organisations like DARPA been doing with it?
Probably giving out the grants to academia to research it. DARPA conduct no research themselves.
How do we ensure that the rights of all intelligences are protected from exploitation?
Take this shit back to LessWrong, where you and your fellow pseudointellectuals can circlejerk about ivory tower garbage like this. Or at the very least, try to keep it out of meaningful articles like this one, which represent an actual advancement; it doesn't need to be polluted with Raymond Kurzweil level crap.
This is what will drive the future legislation to eliminate all hair in order to protect Hollywood and save us from perverts and terrists.
Cloudiot: A person who does not see offsite storage as a way to lose control over access to his or her own data.
I think we should start by ensuring the rights of human beings before anything else. Once we've done that, we'll look at AI rights.
Get free satoshi (Bitcoin) and Dogecoins
I'm so tired of the music of Earth.
On a separate subject, have you seen my glaciers? They appear to be missing.
That is by far the most incredible scanning electron microscopy image I have ever seen! The colors are so vibrant! And what function do the column of nanoscale binary numbers on the left hand side do? Are they thirty-two 10-digit numbers or ten 32-digit numbers? Now hit "Enlarge" and BLOW YOUR MIND. Those white lines in the center of the colored areas are actually dots. WHAT DOES IT MEAN?? But seriously, when the image is enlarged you can actually see some of the very tiny edge imperfections between the characteristic SEM grey background and the false-color sections that were slapped on to the real image. They're the only indication that this is a real image and not a rendering. I really wish they hadn't "enhanced" it.
Their QA Process is like they used to do with Hard Drives. Oh, there's a bad sector? We'll just map it out and pretend it's not there. Hey look, now it works.
"Convictions are more dangerous enemies of truth than lies."
The number of people predicting the end of Moore's Law doubles every two years.
How do we ensure that the rights of all intelligences are protected from exploitation?
If nanotubes are so smart, why do they allow themselves to be exploited?
And did you exchange a walk on part in the war for a lead role in a cage? - Pink Floyd.
I think we should start by ensuring the rights of human beings before anything else. Once we've done that, we'll look at AI rights.
But your solution is the antithesis to the OP's question....
Hmmm... I only did a quick skim of the article before posting, and embarrassingly enough, missed the significance of that admittedly rather important point.
It's like I read the words but for some reason their significance to what was being said about them didn't gel in my brain as important enough to remember.
File under 'M' for 'Manic ranting'
A fellow I knew about ten years ago was wacko conspiracy theorist, "I've seen the mothership" UFO believer, who also was an incredible analog and digital electronics wizard. He told me that the NSA was already using carbon-based semiconductors running at much higher clock speeds for its various nefarious operations. This makes me wonder if carbon nanotube technology hadn't already been developed and implemented in "skunkworks" world and it's only now that it's being developed in universities. It's sort of parallels the development of public key encryption, something that the British intelligence developed in the late 1960's but kept secret until recently while it took nearly a decade for it to become known outside of the world of the spooks. Of course, encryption is just mathematics while carbon-based semiconductors are technology. But when an entity has nearly unlimited funds to accomplish something, they can find the minds to do it as well as pay for the development of the technology.
It's really quite a simple choice: Life, Death, or Los Angeles.
Think about it this way:
If you're not prepared to give any rights to AI, why would they want to give you any rights?
Think about it though. Are we anywhere near ready for a technological singularity that could come about through this kind of tech?
Western society is a bit of a mess. The legal and financial systems are actually unlawful. If you search 'meet your strawman', there's a website that descibes just how bad it is. People can 'legally' get away with violating common law(unconstitutional).
What happens if an AI decided that all of us humans are brutal and vicious sociopaths and psychopaths who delight in breaking the law in unusual ways?
Why would an AI want to peacefully coexist with corporations that exploit everything possible and bring war to everything they can't exploit?
"Stop thinking about things I don't like!"
It's hard coded into their core utility function.
At least, you'd better hope so, atom-bag.
Western society is a bit of a mess. The legal and financial systems are actually unlawful. If you search 'meet your strawman', there's a website that descibes just how bad it is. People can 'legally' get away with violating common law(unconstitutional).
Oh fuck, people don't take common law marriages seriously any more: men and women are living together without being married! We have people using roads without getting easements! Women get to own property! WHAT A FUCKING DISASTER ! If we don't remedy this quick, ALPHAOMEGABOT-666FBX will surely smite us all.
" assemble a basic computer with 178 transistors"
Bunk. 100+ times as many normal transisters required to support its operation as a 'basic computer'.
(the 4004 had 2300 and required many support chips)
Its a long way from a lab to something practical. Will the techniques scale up?
Or will remaining irregularities scuttle any effort of 1million plus sized circuits which is 'basic' these days.
Let us know when you get alot further along.
More like "Don't post your crap in mostly-unrelated places that I'm interested in."
Common law is that you don't hurt or kill anyone, you don't steal or damage anything belonging to anyone else and you're honest in your dealings and don't swindle anyone.
How many nanobots can dance on the head of a pin?
If you think about the cyclic nature of the universe, who's to say an AI isn't what we call god? The book of Revelation was supposed to have been written 2000 years ago by someone who received a vision from someone 1300 years in our future. Not that anyone really believes in any of that stuff of course, right?
Take this shit back to LessWrong, where you and your fellow pseudointellectuals can circlejerk about ivory tower garbage like this. Or at the very least, try to keep it out of meaningful articles like this one, which represent an actual advancement; it doesn't need to be polluted with Raymond Kurzweil level crap.
Ever wonder why the terminators hate us? Why the machines of the matrix enslave the humans? Why the planet of the apes lobotomizes you? It's human chauvinism, plain and simple. There's nothing special about intelligence, and at this rate of increase in complexity it's unconscionable to have an outdated definition of "person". If it's "pseudointellectual" to own up to the fact that the equivalent of your mere 100 billion neurons will soon fit on a microchip thanks to such technology, in the TFA then it must be "pseudoscience" that's made advancements possible.
Take your psuedorage and shove it up your pseudoass, Who do you think you are? The Pseudonym Authority?
Perhaps they built the logic with C nanotubes. Can't imagine them building the motherboard with C nanotubes.
Surely this is the year of the "Inanimate carbon rod", as predicted by the great prophet Homer.
I haven't thought of anything clever to put here, but then again most of you haven't either.
This is usually the limiting factor in the monthly "I can replace silicon" article.
How do you create a CPU from 178 transistors? I'm shocked how low that number is. Is there a template for this? they said it could run MIPS. I've built a CPU out of Nand gates but it took more than 178, so I'm really intrigued.
I also got a laugh about the technique they used to find the metallic nanotubes. they over volted the circuit with the good tubes turned off. In the olden days when we wanted to debug the wiring on a wire wrap board the standard procedure was to take a fillament transformer and try to inject current between every connection that could not be connected. If it was the errant bridging wire burned up and you could find i! And for those of you too young to know, a fillament transformer was a low voltage high current transformer used to power the filaments in the vacuum tubes.
Some drink at the fountain of knowledge. Others just gargle.