³ÉÈË´óƬ

In Nature Photonics, a top journal in optics and photonics, review articles are commissioned by the Editor inviting recognized leaders and pioneers of the field, in this case neuromorphic photonics or photonic computing, to write an authoritative review of the state of the art in the field. 

In the article Photonics for artificial intelligence and neuromorphic computing, Professor Shastri and team not only survey the recent research and current challenges in neuromorphic photonics, but also provide a roadmap for the future of photonic computing by discussing the current and future challenges, and outline the advances in science and technology to meet those challenges.

What is neuromorphic photonics?

Neuromorphic photonics combines photonics (i.e. optical physics) and neural networks (computing inspired by the human brain) to potentially build computers that are powered by light (i.e. photons) instead of electronics (i.e. electrons). Neuromorphic photonic processors could solve problems that are presently very difficult to do so on conventional computers e.g. artificial intelligence, machine learning, solving complex optimization problems  etc.

What are the benefits of Neuromorphic photonics research?

Neuromorphic photonic computing has the potential to revolutionize the speed, energy efficiency and throughput of modern computing platforms. Neuromorphic photonic computing could enable

1. fundamental physics breakthroughs (qubit read-out classification, high-energy-particle collision classification, fusion reactor plasma control);
2. solving nonlinear optimization problems (in robotics, autonomous vehicles) and partial differential equation solving;
3. accelerate machine learning (deep learning inference, ultrafast learning) amongst others

Read the article on ShareIt for the thorough review of the current state and future advances and challenges of neuromorphic photonics.

The article can also be accessed in .

Queen's Gazette's highlight on Prof. Shastri's review article: Computers powered by light and brain networks