Patent award could herald new era in computing

Hewlett-Packard has announced it has been awarded a patent that could remove a major obstacle to making molecular-scale computing a reality.

The patent, issued by the US Patent Office earlier this month to HP Labs scientists Phil Kuekes and Stan Williams, proposes a solution to the problem of connecting molecular-scale devices to today’s much larger integrated circuits.

‘We have a strategy to reinvent the integrated circuit with molecular rather than semiconductor components,’ said HP Fellow Williams, director of quantum science research, HP Labs. ‘We’ve received two key patents and have several more pending that we believe will eventually enable computers to be millions of times more efficient than they are today.’

HP is pursuing molecular electronics as an entirely new technology that could augment silicon-based integrated circuits within the decade and eventually replace traditional solid-state memories. Most experts believe that silicon technology will reach key physical and economic limits by about 2012.

‘Once you’ve built a circuit from molecular-scale devices — something about the size of a bacterium — the question is how you get data into and out of it,’ said Kuekes, computer architect and senior scientist, HP Labs. ‘In order to do that, you have to bridge the size gap between molecular-scale wires and current technology, which is about a hundred times bigger.’

Tiny wires in today’s integrated circuits are addressed through a demultiplexer.

However, building a demultiplexer is said to require an extremely precise, very complex pattern of connections.

Since it would be virtually impossible to make such precise connections with molecular-scale wires, the new patent proposes making connections randomly using a chemical process. The resulting pattern can then be determined using computer algorithms.

‘We’ve essentially created a city of streets crossed by avenues, but they’re so tiny we can’t paint the street signs,’ said Kuekes. ‘Instead, we have a chemical process that gives each street and avenue a unique name. Then we run a program that identifies all the thoroughfares by their names and enables us to create a map of the city. Once you have that map, you can store and retrieve information at any intersection.’

HP and its partners at the University of California, Los Angeles (UCLA), expect to be able to fabricate a 16-kilobit memory using this approach by 2005.

On the web