25 September 2013
Last updated at 22:47
Max Shulaker with Cedric, the first carbon computer: “There is no limit to the tasks it can perform”.
The first computer built entirely with carbon nanotubes has been unveiled, opening the door to a new generation of digital devices.
“Cedric” is only a basic prototype but could be developed into a machine which is smaller, faster and more efficient than today’s silicon models.
Nanotubes have long been touted as the heir to silicon’s throne, but building a working computer has proven awkward.
The breakthrough by Stanford University engineers is published in Nature.
Cedric is the most complex carbon-based electronic system yet realised.
So is it fast? Not at all. It might have been in 1955.
The computer operates on just one bit of information, and can only count to 32.
“In human terms, Cedric can count on his hands and sort the alphabet. But he is, in the full sense of the word, a computer,” says co-author Max Shulaker.
“There is no limit to the tasks it can perform, given enough memory”.
In computing parlance, Cedric is “Turing complete”. In principle, it could be used to solve any computational problem.
It runs a basic operating system which allows it to swap back and forth between two tasks – for instance, counting and sorting numbers.
And unlike previous carbon-based computers, Cedric gets the answer right every time.
“People have been talking about a new era of carbon nanotube electronics, but there have been few demonstrations. Here is the proof,” said Prof Subhasish Mitra, lead author on the study.
The Stanford team hopes their achievement will galvanise efforts to find a commercial successor to silicon chips, which could soon encounter their physical limits.
The transistors in Cedric are built with “imperfection-immune” design
Carbon nanotubes (CNTs) are hollow cylinders composed of a single sheet of carbon atoms.
They have exceptional properties which make them ideal as a semiconductor material for building transistors, the on-off switches at the heart of electronics.
For starters, CNTs are so thin – thousands could fit side-by-side in a human hair – that it takes very little energy to switch them off.
“Think of it as stepping on a garden hose. The thinner the pipe, the easier it is to shut off the flow,” said HS Philip Wong, co-author on the study.
But while single-nanotube transistors have been around for 15 years, no-one had ever put the jigsaw pieces together to make a useful computing device.
So how did the Stanford team succeed where others failed? By overcoming two common bugbears which have bedevilled carbon computing.
First, CNTs do not grow in neat, parallel lines. “When you try and line them up on a wafer, you get a bowl of noodles,” says Mitra.
The Stanford team built chips with CNTs which are 99.5% aligned – and designed a clever algorithm to bypass the remaining 0.5% which are askew.
They also eliminated a second type of imperfection – “metallic” CNTs – a small fraction of which always conduct electricity, instead of acting like semiconductors that can be switched off.
To expunge these rogue elements, the team switched off all the “good” CNTs, then pumped the remaining “bad” ones full of electricity – until they vaporised. The result is a functioning circuit.
The Stanford team call their two-pronged technique “imperfection-immune design”. Its greatest trick? You don’t even have to know where the imperfections lie – you just “zap” the whole thing.
Cedric can only count to 32, but in principle it could count to 32 billion
“These are initial necessary steps in taking carbon nanotubes from the chemistry lab to a real environment,” said Supratik Guha, director of physical sciences for IBM’s Thomas J Watson Research Center.
But hang on – what if, say, Intel, or another chip company, called up and said “I want a billion of these”. Could Cedric be scaled up and factory-produced?
In principle, yes: “There is no roadblock”, says Franz Kreupl, of the Technical University of Munich in Germany.
“If research efforts are focused towards a scaled-up (64-bit) and scaled-down (20-nanometre transistor) version of this computer, we might soon be able to type on one.”
Shrinking the transistors is the next challenge for the Stanford team. At a width of eight microns they are fatter than today’s most advanced silicon chips.
But while it may take a few years to achieve this gold standard, it is now only a matter of time – there is no technological barrier, said Max Shulaker.
“In terms of size, IBM has already demonstrated a nine-nanometre CNT transistor.
“And as for manufacturing, our design is compatible with current industry processes. We used the same tools as Intel, Samsung or whoever.
“So the billions of dollars invested into silicon has not been wasted, and can be applied for CNTs.”
For 40 years we have been predicting the end of silicon. Perhaps that end is now in sight.