Slashdot Mirror
Nanotech in Microchips by 2015
dotwhynot writes "Molecular electronics, a realm once considered science fiction, could be heading for our computers and devices sooner than thought.
A new report on the technology roadmap of the chip industry finds a growing confidence in new nanotechnology, and forecasts that the transition to the post-silicon era could happen by 2015.
The development of nanoswitches has already reached a point where it will be possible to manufacture them reliably at low cost. Intels goal over the next decade is to build chips that hold more than one trillion switches."
The transition to new nanotechnology techniques could occur around 2015, when chip makers will have exhausted their ability to shrink the wires and switches that make up the modern processors and memory storage devices at the heart of the computer, communications and consumer electronics industries.
Nevermind the growing heat concern. Who was it that said soon microchips will be hotter than the surface of the sun if they keep getting faster at the same rate they are now?
Bradley Holt
And there was me thinking that microchips manufactured on the 65nm scale was nanotech.
I have never let my schooling interfere with my education. - Mark Twain
Well, it's actually a hoax:
http://www.snopes.com/inboxer/hoaxes/computer.asp
Although the photograph displayed could represent what some people in the early 1950s contemplated a "home computer" might look like (based on the technology of the day), it isn't, as the accompanying text claims, a RAND Corporation illustration from 1954 of a prototype "home computer." The picture is actually an entry submitted to a Fark.com image modification competition, taken from an original photo of a submarine maneuvering room console found on U.S. Navy web site, converted to grayscale, and modified to replace a modern display panel and TV screen with pictures of a decades-old teletype/printer and television (as well as to add the gray-suited man to the left-hand side of the photo
First off, there is exactly 0 reason to switch from QWERT to Dvorak. The only proof that Dvorak is faster is from..oh what was that guys name? hmm oh yeah, Dvorak.
QWERTY has nothing to do with speed, and everything to do with the letter positioning in the carriage of typewriters.
Even the alledged speed difference is pretty much moot on any modern computer, for all practical reasons.
Now, there is hugh motivation to make this technology work. When it does work it will mean faster smaller and cooler computers. That means big money to Intel, and also to AMD who will wait for intel to develop it, then steal it...then possibly make it better.
they won't convert fabs, they will build new. And spending a billionh dollarsd an a fab so you can make 100 billion is a good investment, espcially if you control the IP.
The Kruger Dunning explains most post on
To help you out, there are three levels of circuit design you would need to learn. The first is the basics. What the NAND, NOR, OR, AND, XOR, and various flip-flops do. That was a sophmore level course in computer/electrical engineering (or CS) at my school. After that, you get into more advanced designs including designing your own microprocessors and such. That level is the most advanced you can learn without being "process specific".
The last level is the physical level. Currently, this involves laying out the parts of the transistors on circuit, all the metal, all the N and P areas (If you don't know what that means, you will when you take basic electronics courses involving transistors) and so on. With this new tech, you will have to relearn this area, but that is it.
Fly me to the moon Let me sing among those stars Let me see what spring is like On jupiter and mars
There are a lot of issues with carbon nanotube (CNT) technology that your post doesn't take into account. Carbon nanotube transistors already exist in FET form, but there are hurdles to overcome in terms of integrated circuit production.
The first thing that will need to occur is selective growth of semiconducting or metallic carbon nanotubes. All the current synthesis methods that I'm aware of produce a hodgepodge of both, and separating them is not the easiest thing to do.
The techniques of which I'm aware for device fabrication using CNTs tend to use some variation of flowing a gas such as methane over an iron-based catalyst at high temperature (~900 deg. C). The result is nanotubes growing in every direction from the chunk of catalyst, and using lithographic techniques you can then attach electrical leads to the islands and hope that they're connected by a single tube. This is quite successful for research purposes, but needs some serious optimization/automation for commercial fabrication.
I think what I'm trying to say is that your comment that current technologies will be "optimized for nanotube wires" overlooks that fact that conventional technology is not necessary appropriate for such adaptation.
We're still a long way from integrated CNT transistor circuits.