**By combating the inherent limitations of electrons, optical computing uses photons rather than electrons to perform calculations, which can significantly improve the speed and energy performance of calculations.**

The engagement of light and matter is the fundamental principle of optical computing. Matrix computing has become one of the most widely used and essential information processing tools in science and engineering, contributing to a large number of computational tasks for most signal processing, such as discrete Fourier transforms and convolution operations.

Matrix multiplication consumes the majority of computational resources as the fundamental building block of artificial neural networks (ANN). Simple matrix multiplications involve a large number of transistors collaborating together due to the characteristics of electronic components, while matrix multiplications can be easily applied by basic photonic components such as micro-ring, Mach Zehnder interferometer (MZI) and a diffractive plane.

As a result, optical computing is several orders of magnitude faster than electronic computing and uses far less power. The conventional technique of incoherent matrix-vector multiplication, on the other hand, is limited to real-valued processes and does not work well with complex-valued neural networks or discrete Fourier transforms.

Huazhong University of Science and Technology (HUST) scientists led by Prof. Jianji Dong presented a photonic complex matrix-vector multiplier chip that can assist arbitrary large-scale, complex-valued matrix multiplication operations.

The chip overcomes the bottleneck that conventional non-coherent optical computer schemes face when performing arbitrary large-scale complex-valued matrix multiplications, while enabling artificial intelligence implementations such as Discrete Fourier Transform , the discrete cosine transform, the Walsh transform and the image convolution.

Their plan is to create smart matrix decomposition and matrix partitioning algorithms for micro-ring network architecture to extend matrix multiplications from real to complex domains and from small to large scale.

Scientists have conclusively demonstrated many typical applications of artificial intelligence, illustrating the promising applications of the photonic complex matrix-vector multiplier chip in artificial intelligence computing.

**Journal reference:**

Chen, J. *et al*. (2022) A small micro-ring array that performs large complex-valued matrix-vector multiplication. *Frontiers of optoelectronics*. doi.org/10.1007/s12200-022-00009-4.

Source: http://english.hust.edu.cn/