The 4 computer systems of the future (and what we will use them for)
Thirty years ago, the height of chic was to own a portable landline phone. Now our smartphones are personal computers capable of processing natural language commands and running AI models on the device.
In 30 years, experts say, we will have flying cars, robot butlers and colonies on Mars. Law?
Maybe, maybe not. The next 30 years of computer advancement do not seem as certain as the previous ones.
We are moving away from Moore’s Law and starting to see diminishing returns when it comes to building more powerful classical systems.
On the other hand, we are also on the cusp of several new computing paradigms. And it’s clear that at some point we’ll be moving beyond traditional supercomputing.
Whether that happens in the next 30, 50, or 100 years, however, is a different question.
So, for the sake of conversation, let’s imagine a world where the four major computing systems of the future (classical, photonics, hybrid, and quantum) have all become useful. We will be optimistic and say this happens within the year 2051.
I have bad news for anyone considering shortening binary systems in favor of emerging quantum technology: Classical computers are going to be around much longer than the next 30 years.
Binary computers are to quantum computers what pen and paper are to hadron colliders. The average person won’t need direct access to a quantum computer or hadron collider in their lifetime, but we all benefit from their existence.
That being said, think about your iPhone. You will still have something similar 30 years from now. It might be glasses or a neural implant (unlikely, but possible). Either way, it will need enough on-board processing power to run discrete algorithms and applications. And that’s something an iPhone can do today.
And, just like today’s smartphones, those of the future must above all be powerful enough to connect to the cloud.
Binary computers, in the future, will do most of the same things they do now. And for tasks requiring more power than can reasonably be expected from a future PC, they will still function as an interface to more powerful systems.
It is exciting. Photonic computer systems don’t quite exist yet, but the big idea is to use photons to do calculations instead of electricity. Electrons can’t travel so fast as photons travel at the speed of light because, you know, they’re literally are light.
This means that it is (theoretically) possible to create a computer system capable of processing information at the speed of light.
Researchers at IBM and the Institute of Science and Technology in Skolkovo recently developed a functional photonic switch, a device that could essentially replace silicon-based transistors.
Photonic computers could be thousands of times faster than today’s most powerful binary supercomputers, and because of how they work, they would actually require less power to operate.
It is entirely plausible that this technology could mature over the next thirty years and the biggest benefit that we can all see then is the advent of level five autonomous vehicles.
Yes, level 5 is the big one. At the maximum level of range, a vehicle could operate on its own entirely off the grid and without human supervision.
This would be made possible, in essence, by inserting a gigantic supercomputer into a small automobile. But instead of “giant supercomputer”, we substitute “small photonic computer” and assume that it uses 1 / 100th of the energy while producing 100 to 1000 times the power of its classic cousin.
Here we are specifically referring to classical-quantum hybrid systems. It is possible that photonic computers associate well with quantum systems, but it is beyond the scope of this article to engage in double-blind speculation.
We mentioned earlier that all quantum systems are likely to require some form of classical system to function as a portal, interface, or controller. But there is also a paradigm in which a system switches between classical and quantum calculations or combines the results of both in order to run specific algorithms.
What is interesting here is that these systems are likely to be the first commercially purchased “quantum computers”. Keep in mind that we are unlikely to solve quantum computing to such a degree that you can install a working time travel processor in your basement within the next 30 years.
But that doesn’t mean there won’t be a happy medium. Quantum systems are targeted solutions to very specific problems. You can’t just install an API on the IBM Q system and tap into quantum worms to speed up video rendering, for example.
But you could, theoretically, build a system that runs an airport’s flight planning software through a combination of classic multitasking (to manage infrastructure) and quantum algorithms (for a mathematical plot that’s too complex for a processor. traditional).
Given that similar systems already exist in rudimentary form, it is evident that the next 30 years will see large companies (those with around a billion in value) purchasing and installing hybrid quantum systems as the basis of their computing stack.
Here is the fun part! Quantum computers are about 20 years away, and depending on who you talk to, they still could be.
Quantum computers today are experiments that are painstakingly built in huge labs at great expense, and in essence they solve a big math puzzle or two. It is simply impossible to guess when a functioning and useful quantum computing system will arrive.
But it is quite possible that this will change in the next 30 years. And, if it does, the rest of the science and tech world will too.
Really useful quantum computers could help us solve cold fusion, warp drives, and artificial intelligence in general.
It’s hard to overstate the potential of quantum computers. The implications for the fields of chemistry, drug discovery and pathology alone are incalculable. Billions of lives could be saved as thousands of diseases are eradicated by science.
But when it comes to exploiting spooky actions from a distance or calculating at the speed of light, the future is uncertain. It could take 10, 30, or even 100 years for any of these technologies to mature.