Advanced Chip Tech Using Light Instead of Electricity Could Change Everything

This technology is also expected to drastically reduce the amount of energy consumed by such computers, which currently devour high amounts of electrical power as an unfortunate side effect of running fast.

The new design utilizes a silicon-photonic (SiPh) chip, an entirely new kind of computational semiconductor innovation.

Based on research by Professor Dr. Nader Engheta of the University of Pennsylvania School of Engineering and Applied Science, it makes use of methods Engheta’s team developed to manipulate nanoscale materials so they can be used to carry out math functions normally reserved for conventional semiconductor chips using light rather than electricity to power and communicate the mathematical results. The new chips also still use silicon, the same inexpensive core material used in conventional semiconductor devices. The similarity in tech could also mean the transition to commercialization for this entirely new kind of chip could be shorter than with other technical breakthroughs of similar potential.

The idea which led to the current breakthrough was a merger of two separate innovations. One came from Engheta’s concepts involving how to use conventional silicon tech using light rather than electrical transfer to perform math calculations. The other came from University of Pennsylvania Associate Professor Firooz Aflatouni, whose team developed breakthrough techniques involving nanoscale devices of various kinds.

“We decided to join forces,” Dr. Engheta said about their collaboration.

In partnering on this innovation, Engheta and Aflatouni were after more than just computational speed. They also had the objective of creating a platform to carry out what is known as vector-matrix multiplication. That is a core mathematical operation used in the development and function of neural networks, the basic computer architecture which provides the backbone for today’s modern artificial intelligence (AI) networks.

According to Engheta, one of the core aspects of the SIPh chips involves creating a silicon wafer of non-uniform thickness, as opposed to the uniform ones used in the semiconductor industry. And while the process does involve making the “silicon thinner, say 150 nanometers”, Engheta said, the thickness will vary across the chip.

Those variations in height but without the addition of any other materials, the scientist explained, provide a means of controlling the propagation of light through the chip, since the variations in height can be distributed to cause light to scatter in specific patterns, allowing the chip to perform mathematical calculations at the speed of light.

Aflatouni says that the current design is already ready for commercial applications, and could potentially be adapted for use in graphics processing units (GPUs), the demand for which has skyrocketed with the widespread interest in developing new AI systems. “They can adopt the Silicon Photonics platform as an add-on,” says Aflatouni, “and then you could speed up training and classification.”

In addition to faster speed and less energy consumption, Engheta and Aflatouni’s chip has privacy advantages: because many computations can happen simultaneously, there will be no need to store sensitive information in a computer’s working memory, rendering a future computer powered by such technology virtually unhackable. “No one can hack into a non-existing memory to access your information,” says Aflatouni.

This study was conducted at the University of Pennsylvania School of Engineering and Applied science and supported in part by grants from the U.S. military. 

The research describing how the new SiPh was created and how it performed, “Inverse-designed low-index-contrast structures on a silicon photonics platform for vector–matrix multiplication,” by Vahid Nikkah, Ali Pirmoradi, Farshi Astiani, Brian Edwards, Firooz Aflatouni, and Nader Engheta, all from the University of Pennsylvania School of Engineering and Applied Science, was published in the February 16, 2024, issue of Nature Photonics.

Publishers Commentary

Commercially available photonic chips means that AI of all types will become far more powerful and affordable. It won't just be massive tech companies who produce strong AI but anyone with a bit of money and programming skills. 

With that strong AI will come the rapid weaponization, criminalization and misuse of strong AI. Those not using equally strong AI to protect their computer systems will become easy prey. Some legacy systems will have to be taken offline or use other something other than the Internet for communication. 

It will also mean that the enormous amounts of energy consumed for cryptocurrency will be greatly reduced and that blockchain technology can spread further, but will also be easier to compromise. 

The rush to replace human creativity with AI will accelerate in media and the quality of AI generated media will surpass that of human created media. It will massively disrupt the film, TV, book and music industries. It will also disrupt spectator sports and result in virtual blood sports. The real and synthetic will become impossible to distinguish at some point with the synthetic becoming more appealing because it will be more perfectly suited to consumer's needs and desires. AI will also enable media to more fully integrate mind control and to massively increase the effectiveness of that mind control and social engineering.

If not constrained, AI on photonic chips will also mean great technological leaps forward and the ability to solve major problems.

It is not yet known if the U.S. government will try to control the manufacture and distribution of photonic chips or not.