An EPSRC-funded project at the University of Bath aims to develop technology that could one day form the processor for the next generation of super-computer.
These photonic computers — devices run using light rather than electronics to convey information — would be a million times faster than current silicon-based computers, it is claimed.
Their development would allow computers to evolve beyond the limits of current electronics. The development of ever-smaller silicon chips has allowed computers to double their processing power every 18 months, as detailed by Moore's Law. But this process will one day come to a halt, as once shrunk below a certain size, silicon chips cannot function properly.
Physicists led by Dr Fetah Benabid will look at developing attosecond technology — the ability to send out light in a continuous series of pulses that last only an attosecond, or one billion-billionth of a second.
Starting in June the research, which will receive an £820,000 grant, will develop a new technique enabling the team to look at the temporal profile of light and synthesise its waveform.
Synthesise waveform
'Light in the ultraviolet, visible and infrared range oscillates very quickly, at a rate of 10^14 or 10^15 per second,' said Benabid. 'But if you can capture this oscillation then the problem is how to synthesise the desirable waveform necessary to process information.
'To do that, you need a lot of frequencies over a broad spectrum; in other words, a lot of colours of light must be used to be able to process information.
'The two factors of temporal profile, or waveform, are directly connected to the profile of this ultra-broad coherent spectrum and you can change the temporal profile by diminishing the magnitude of certain frequencies of light that are being produced, altering the waveform.'
Waveform synthesis is the ability to control precisely the way electric fields vary their energy. This control of the variation of the electric field will allow devices such as computers to function in the precise way needed.
The team will build a synthesiser that will form the equivalent of 40 to 50 lasers to create the attosecond light pulses and arbitrary waveforms, which could be used for computing or other applications.
However, using lasers is also problematic. 'You would need to use a lot of lasers at once to move information,' said Benabid.
'This would be cumbersome and with a laser you can lose a number of frequencies over the broad spectrum of light — this use requires all the colours to be coherent, and our work is to make this possible.'
The researchers will use photonic crystal fibres which, unlike conventional optical fibres, can channel light without losing much of its energy.
In the research, light of one wavelength will be passed down a photonic crystal fibre, which then branches off in a tree-like arrangement of fibres, each with a slightly separate wavelength, creating a broad comb-like spectrum of light from ultraviolet to the middle of the infrared range.
Close control
This broad spectrum would allow close control over the electric field, which is the basis of conveying enormous amounts of information that devices such as computers need.
'By merging our research with quantum computing we could take a step closer to making quantum computers a reality,' said Benabid. 'Quantum researchers would have to address the issue of data storage but this work could form the core of the computer's processor.'
He added that if his work was successful it could lead to the development of lasers that operate at wavelengths current technology cannot now create, which would be important for surgery.
The continual series of short bursts of light could also advance physics by giving researchers the ability to look inside the atom and examine its dynamics.
By sending the light in short bursts into an atom, it may be possible to work out the movements of electrons as they orbit the atom's nucleus and may also reveal more about sub-atomic particles.
Telecommunications could also benefit, as the technology would allow large increases in the rate of data able to be transferred.
Record set at EPFL hyperloop test facility
The problem with hyperloops is capacity. High speed and small vehicles limit capacity. Junction design is key. Road vehicles can change lane with...