Until now quantum computers have competed with classic computers in solving problems. But, researchers are planning to leverage the power of quantum computing for making a neural network that acts like a brain.
The Neuromorphic Quantum Computing (Quromorphic) project led by Michael Hartmann and his colleagues from Heriot-Watt University is all about combining two dominant technologies in today’s world — neural networks and quantum computing. To mimic the brain, scientists use AI models that are trained beforehand using a set of data. However, Hartmann and his team want to focus on the hardware aspect of the brain.
“My colleagues and I instead hope to build the first dedicated neural network computer, using the latest “quantum” technology rather than AI software.”
One of the major advantages of using quantum technology is the number of inputs it can take in the form of qubits. While traditional computers store data in bits which can accept either 0 or 1 as input, quantum computers, with qubits, double the computing power. This enables such computers to process a huge amount of data simultaneously.
Hartmann says there are plenty of technical challenges involved in the project, but AI has a high threshold for tolerating errors as compared to conventional computers. Therefore, the machine won’t “need to be quite so well isolated.”
To develop a quantum neural network and put the technology inside a device to harness its power, it might take ten or more years as the researchers need to control technical details to avoid computational errors. However, once developed, it will mark the beginning of a new era in technology.