CreativeWire

In what can be considered a crucial step towards reaching the ultima Thule of computing, scientists from Oxford, Münster and Exeter Universities, have developed microships that mimic the the human brain by combining computing and memory into a single unit.

Conventional computers based on the von Neumann architecture process information by splitting memory and computing into separate units. While this architecture works well for typical computing tasks, its limitations become apparent when dealing with big data processing and artificial intelligence. That is why developing computing systems that mimic neuro-biological architectures has become a necessity, since synapses in the human brain act both as the processing and the memory unit, thus allowing the brain to be energy efficient, all the while being able to process a wide range of information and emotions. 

The research team led by professor Harish Bhaskaran from Oxford University, have created a photonic microship that behaves like the brain's synapses, by exploiting the fact that synapses are by far the most numerous constituent of real brains, outnumbering neurons by several orders of magnitude. The ship not only exhibits brain-like behavior, but also takes advantage of its photonic nature - by using light rather than electricity - to deliver signals between its artificial neurons. This design allows the microship to be energy efficient, while at the same time being able to process information at speeds that exceed those observed in the human brain.

"The development of computers that work more like the human brain has been a holy grail of scientists for decades. Via a network of neurons and synapses the brain can process and store vast amounts of information simultaneously, using only a few tens of Watts of power. Conventional computers can't come close to this sort of performance.", said the lead researcher.

The research was published in Science Advances on Wednesday, September 27 2017. The researchers believe that their study could pave the way for a new era of computing, effectively laying the foundation for building efficient neuromorphic systems that can one day think and process information just like a typical human being.