Neural networks will switch to optical chips
Stanford University researchers have found that training artificial neural networks directly on an optical chip leads to more energy efficient execution of complex tasks, such as speech or image recognition..
An artificial neural network is a type of artificial intelligence that uses connected units to process information in a manner similar to how our brains do. Optical-oriented devices are of great interest because they can perform computations in parallel using less power than electronic ones. Researchers have developed an optical chip that trains neural networks in the same way as conventional computers.
In the new teaching protocol, the laser is fed through an optical circuit. On exiting the device, the difference from the expected result is calculated. This information is then used to generate a new light signal that is sent back through the optical network in the opposite direction. The phase regulator settings can be changed based on this information, and the process can be repeated until the neural network produces the desired result..
Researchers are planning further optimization of the system and want to use it to solve practical problems. The general approach they developed can be used with various neural network architectures and for other areas such as reconfigurable optics..
Making A Neural Network Using Light? | Optical Neural Networks, Explained
Recall that the Pentagon began to create a military command system with the help of thoughts.
text: Ivan Malichenko, photo: Stanford University