(By Kit Eaton)
“What Stanford has done goes far beyond replicating a logic gate. The team managed to create 178 carbon nanotube transistors and wire them together to become a very basic processor. To prove their success, they then popped a basic operating system into the computer that ran on the famous MIPS basic instruction set, and managed to program it to do tasks like number sorting and counting. Though that sounds pretty simplistic, the fact they assembled 178 breakthrough-tech transistors in the first place is astonishing.“
We just wrote about the advances in post-silicon chip technology made by Northwestern University, and now Stanford University has revealed it’s made an entire computer out of chips that aren’t based on silicon–they’re based on carbon instead.
Specifically the Stanford advance is based on carbon nanotube transistors which, as we learned from the NU research, can operate at astonishingly low power levels. This trick gets around one of the biggest problems in even the most sophisticated silicon tech, which is the fact that cramming billions of transistors needed for a CPU onto a small chip die leads to a lot of heat generation. This is simply a side effect of the electrical energy flowing through the transistors, and it has all sorts of important implications. The most obvious of these is that really powerful chips can need huge heat sinks, and that mobile chips can’t necessarily run at full power because they’ll get too hot for being inside a smartphone. The lower power demands of carbon transistors means you can squeeze many more of them into a chip without causing overheating, or you can use the same number of transistors but run your chip on just a fraction of the power.
What Stanford has done goes far beyond replicating a logic gate. The team managed to create 178 carbon nanotube transistors and wire them together to become a very basic processor. To prove their success, they then popped a basic operating system into the computer that ran on the famous MIPS basic instruction set, and managed to program it to do tasks like number sorting and counting. Though that sounds pretty simplistic, the fact they assembled 178 breakthrough-tech transistors in the first place is astonishing.
The real success of this research, of course, is in the hard science needed to build the carbon nanotube transistors. These rely on neat parallel lines of nanotubes being grown on a chip, but it’s tricky to make them grow like this–and even a 99.5% alignment rate wouldn’t be good enough on a chip containing billions of transistors because that would lead to misbehaving circuits. Secondly, as part of the growth process some nanotubes can end up acting as tiny conducting metal wires, which is absolutely not what you want in a semiconductor chip. So Stanford’s scientists worked out how to evaporate these failed tubes into a puff of carbon dioxide, leaving only the right kind of material behind to become transistors. To avoid the misaligned tubes, the team worked out an algorithm that could work on any old array of nanotubes even if some of them are actually askew.
Both of these advances are exactly the sort of innovation you’d need as part of a mass production run of carbon nanotube chips. This doesn’t mean we’ll see this chip tech in 2014’s iPad though, as it’ll take years to mature. But it does mean that near future computers may deliver phenomenal power and battery life thanks to one of the basic elements that also makes up life on Earth.
“Opinion pieces of this sort published on RISE Networks are those of the original authors and do not in anyway represent the thoughts, beliefs and ideas of RISE Networks.”