Soon, computer made from 'invisibility-cloak' materials?

Scientists have found that materials which make an invisibility cloak a real-life possibility can perform advanced mathematical calculations, paving way for development of a new kind of analogue computer.
An international team of researchers has found that so-called metamaterials, which can alter the properties of light waves often to render an object invisible, could perform mathematical operations as well.
Nader Engheta, at the University of Pennsylvania, and colleagues decided to explore a different use for metamaterials, one that adapts the old idea of analogue computing.
Analogue computers were limited in precision by the materials available at the time - for example, anything requiring moving parts for computation was limited by how small those parts could be made.

But metamaterials, which rely mainly on light, have no such constraints. Engheta's team has simulated a metamaterial capable of calculus functions like differentiation and integration, and other fundamental mathematical tools, 'New Scientist' reported.
The metamaterial computer works because light waves can draw mathematical curves in space, akin to a graph. In calculus, differentiation describes the slope of that curve at various points, while integration gives the area under the curve.

The team's metamaterial block can perform these calculations by modifying the light wave's profile.
"As the wave goes through this block, its profile changes such that by the time it comes out it has the profile expected from the given mathematical operation," said Engheta.

So far the team had only simulated these metamaterials, as both an ideal theoretical material and a more realistic one made from precisely alternating layers of silicon and aluminium-doped zinc oxide.
"As our near-future action item, we are planning to build our proposed blocks in order to test the proof of the concept," said Engheta.
If it works, the new metamaterial could in theory be included in camera lenses to perform image processing tasks like edge-detection or pattern recognition, which is useful for identifying faces and other objects in pictures.
At the moment such tasks are done pixel-by-pixel using an ordinary computer chip, but a metamaterial computer could process the entire image at once.

The study is published in the journal Science.