Quantum Computing Is Old. Meet Photon Computing.
Beyond quantum computing: explore how light-based photonic chips could power the future of data centers.
Once, a wise man—probably one of the wisest ever lived on the earth—said that nothing (literally nothing) in the universe can travel faster than light. They often give the example that even travelling at 1% of the speed of light gives rise to mind-bending phenomena such as the Twin Paradox.
But what will happen if our computers start processing information at—or at least closer to—the speed of light?
In the previous article on quantum computing, we discussed photon computing and how some scientists believe it could eventually overcome several limitations of conventional and even quantum computing for certain applications.
While the world is still discovering the astonishing capabilities of quantum computers, the scientists and the technology companies have already started working on another potentially transformative technology: Photon Computing.
But what exactly is Photon Computing? And before that, what is Photon?
The Light Particle, Literally Light
As explained in the previous article, at extremely small scales, things are both substance (particles) and waves. Light is no exception. It behaves both as a wave and as a particle, and the particles of light are called photons.
In other words, just as the things we touch or hold are made up of particles such as electrons, protons, and neutrons, light is made up of photons—which have no mass (meaning they weigh nothing).
It may sound mind-bending to those learning this for the first time, but that’s simply the way it is.
Now, almost all computers, laptops, smartphones, servers, and supercomputers built to date run on electricity. Anyone who has read even a little about electricity would know that electricity means the flow of electrons. Similarly, the flow of light particles (photons), and computing based on that flow, is what we call Photon Computing.
The Heat Problem and The "Light" Solution
But the benefit of Photon Computing is not only the speed. You may have heard that, due to the increasing use of AI, data centers generate a lot of heat. And a lot of water is required to keep these data centers cool.
This could become an even bigger problem in developing countries like India, where large technology companies are building data centers in areas that already face water shortages. As data-center infrastructure continues to expand, the pressure on water resources could increase further.
A possible solution to this problem could be Photon Computing, because much less heat is generated when processing information with light.
Less Heat → Less Cooling → Less Electricity → Lower Operating Costs.
This is one of the reasons why some of the world's biggest technology companies are investing billions of dollars in photonic computing and related technologies.
Now think about it. As rapidly as AI is advancing today, the number of data centers required to support it is also increasing. Each new generation of AI models demands more computing power than the previous one.
More Computing Power → More Servers → More Electricity → More Heat.
Some experts even believe that the biggest problem AI could face in the future may not be a lack of computing power, but a lack of electricity. And Photon Computing is being considered as one possible solution to this problem.
The Solution Alredy Existing
Let me tell you something interesting: we have been using light for communication for decades. You must have heard of optical fiber. Today, much of the world's Internet runs through millions of kilometers of optical fiber cables, where data already travels in the form of light.
Scientists are now trying to bring this technology inside computers. In other words, the same light that is carrying data between cities and countries today could potentially process data inside computer processors tomorrow.
Think about it: the technology that helps run the Internet today could one day be working inside your computer's processor as well.
So how does the Photon Computing actually work?
Bringing The Idea Into Reality
In the chips used in today's computers, electricity keeps flowing through tiny transistors and thin copper wires. But in a photonic chip, instead of electricity, light generated by lasers is sent through extremely fine optical paths.
Different optical components built into the chip then bend and control that light, and data processing takes place based on how the light behaves. Such a chip is called a Photonic Integrated Circuit, or PIC.
Well, silicon does not have to be replaced immediately in photonic chips. Rather, photonic chips can be built using silicon itself. And that is why this field is called Silicon Photonics.
Many of the world's biggest technology companies are working in this direction. Intel has been researching silicon photonics for many years. IBM is also working on this technology. NVIDIA is investing in optical networking for faster AI systems.
Apart from these, many startup companies such as Lightmatter, Lightelligence, and Ayar Labs are also trying to turn photonic computing into a commercial reality.
But this does not mean at all that a photon processor will come to your laptop next year. This technology is still in its early stages.
Many Miles To Go
Just as the first-generation computers of the 1940s and 1950s were built using vacuum tubes, a decade or two later came the second-generation computers with magnetic-core memory. Then came third-generation computers built with transistors. Around 1970, microprocessors made using silicon chips emerged, and they are still in use today.
Similarly, it is possible that we are now entering the beginning of a new era of photon computing.
Vs. Quantum Computing
In the previous article, I talked about Quantum Computing. But Quantum Computing is not going to replace our laptops and desktops. It is meant for certain types of highly complex problems.
If any major revolution is going to come to the world of laptops, desktops, and data centers in the future, then Photon Computing could be one of its strongest contenders.