Slashdot Mirror
First Ever Nanotube Transistors On A Circuit
btsdev writes "Researchers at the University of California at Berkeley and Stanford University have developed the first ever integrated silicon circuit with nanotube technology. According to the article on UC Berkeley's site, this brings researchers one step closer to developing memory chips with carbon nanotubes - chips that could hold approximately 10,000 times more data than those we have today."
Berkley has made some great stuff over the years. But this is truly cool. You could make a supercomputer the size of your current computer tower today. Or maybe even smaller with some other control method.
Or even maybe implant it in your body.
If you could get lots of small chips to give high memory density, pack them into a PC and then setup a huge RAM disk with some permanent storage things would suddenly become a lot faster
Rus
CPanel + Root from $35/mo - 10% off with discount code SLASHDOT
Ok, if you have 10,000 more the space, it all disappears when you power off right? Or when the power goes out?
Also what about address space?
How many bit CPUs will we need to address 1,280,000MB of RAM?
Nonetheless cool, even though it seems either overkill or impractical
Error 407 - No creative sig found
carbon nanotubes - chips that could hold approximately 10,000 times more data than those we have today
How about carbon Megatubes that could hold (10x9)* (10x6) times the data of carbon nanotubes
Diamond substrates and nanotubes face completely different challenges, and the issues with nanotubes will probably be resolved first. In that WIRED article, it was explained that it takes years to grow ONE diamond wafer, and they still haven't probably grown anything larger than 3-4 inch wafers. It will probably take several decades until they can serially produce 12" diamond wafers.
Carbon nanotubes, on the other hand, need to have their type (metallic or semiconducor) and doping level (if semiconductor) controlled reliably, and also EDA tools extentions need to be written in order to incorporate them in critical paths on chip. This is a lot less work than learning to grow diamond wafers. At least De Beers will need to be brought to its knees before this will be reality. Imagine a war over diamonds...
At least in a server environment, I don't see the requirement for many gigs of memory (on a single chip no less) without also having better technology to access it quickly.
Ok.. Now imagine those many gigs of memory on on-die with the CPU itself. Get's interesting, yes?
Trolling is a art,
Actually, the idea of building "integrated vaccum tubes" isn't as silly as it sounds. Transistors don't function above 200C, and microscopic tubes would allow us to build sensors and other circuits where transistors cannot go, at least without elaborate cooling. There has already been talk of using silicon vaccum tubes to power remote sensors in jet and aircraft engines, which must operate at extremely high temperatures.
And I always thought they would find an idea home in robot spacecraft, where there is already a vaccum. They would also offer extreme resistance to the effects of hard radiation such as the Io belt around Jupiter, which tends to fry semiconductor electonics.
My rights don't need management.
They actually won't have 4-inch wafers for another few years - I think they're probably at about an inch squared right now. It won't take several decades to produce 12" wafers, though, because the size of the wafer they produce using CVD depends on the size of the seed. They are using the result from each session to grow larger wafers each time: Starting with a square, waferlike fragment, the Linares process will grow the diamond into a prismatic shape, with the top slightly wider than the base.
Still, you're right - nanotubes will probably be feasible before mass-produced diamond ICs. The reason I brought up diamond/nanotubes is because they're both carbon - two very different forms of it - and I'm kind of interested in how the process of trying to get one onto the other would affect both of them.
I claim first use of "Error No. 0B" - or "No. 0B error." It'll be the new ID 10T!
And people these days think that Fossil Fuels are the result of a few million years of pressure and heat transforming Dead Trees.
In fact all these "fossil fuels" we keep burning are the decomposition of a once well-known and essentially pervasive vastly superior technology. Technology which we're only now beginning to open the doors to.
Visit CryptoGnome in his home.
My prediction is that the first high-tech consumer product implants will be cell phones. But this does raise interesting questions about producing reasonably sized implant electronics for blind and deaf people, as well as other human systems failures.
"Who are in control, they are not in control of anything - they don't even control themselves!" - Glen Beck
until they can encode the human genome in something close to the size of the human genome?
Information on the Caltech research can be found here.
I seem to remember that Nantero had a 10Gb working prototype for NRAM. Is this significantly different from what they are doing?
http://www.nantero.com
Reading the article, it looks like what they did was build a chip that can detect the types of nanotubes growing on it - conductive or semiconductive, with the nanotube actually being grown on the chip itself.
This research is a nanotube manufacturing method, not nanotube circuit fabrication.
-R
I'm surprised that the Berkeley people grew the tubes on the semiconducting substrate (and skeptical that this is the way to go). Unless I am misreading the article (always a possibility), they have created a very expensive way to evaluate only thousands or millions of tubes per manufacturing cycle. I would think that the real key to low-cost nanotube circuits is to use bulk chemical processes.
Using bulk chemical processes, one might grow a big batch of nanotubes, harvest them, sort them by size (centrifuge), chemically modify them to have certain electronic properties (i.e., attach functional groups to the surfaces or tube ends), and sort them electrochemically (perhaps with eletrophoresis). I can envision any number of interesting bulk chemical processes that simultaneously modify, test, or sort nanotubes. These bulk processes would yield batches of trillions or quadrillions of near-uniform, high-quality semiconducting nanotubes with each cycle of the process.
And instead of using lithographic techniques (printing an accurate pattern of circuits on a wafer), I would expect nanotube circuits to be chemically deposited using self-organizing chemical films. These self-structured films can have feature dimensions far smaller than anything semiconductor maker can fabricate. The only need for lithography is at the edges -- creating an interface between the macroscopic off-chip interconnects and the nanoscopic fields of nanotube memory zones.
Two wrongs don't make a right, but three lefts do.
This article has the researchers at Berkeley claiming to be the "first ever" to report success in integrating nanotubes with integrated circuits. What about that company Nantero which claims a propriety nanotube memory chip design ( NRAM ), developed by Dr. Thomas Rueckes (who got his PhD in chemistry from Harvard).They have venture capital ( from Charles River Ventures, Draper Fisher Jurvetson, Stata Venture Partners, and Harris & Harris group). Their web page (www.nantero.com) claims, " Dr. Rueckes' pioneering design takes advantage of these unique properties while cleverly integrating nanotubes with traditional semiconductor technologies for immediate manufacturability." This makes it sound like they may have a product "real soon now". Are the guys at Berkeley not aware of the work done by these "Harvard guys"? Is this an "East Coast" vs. "West Coast" rivalry? or is it just academics not being aware of what's going on commercially? or is Nantero trying to "pull a fast one" and really aren't as far along in developing "NRAM" as they imply?