## Traditional Methods

Traditional storage means a lot of things, right now. Magnetic storage is still used pretty consistently, as SSDs aren’t quite at the point where they replace everything like hard drives themselves did.

Now, quantum computing occasionally hits the news when a major breakthrough happens, and for good reason! Quantum computing promises to do more than any major storage advancement before. Quantum computing *isn’t *just ‘better’ classical computing – it’s a whole new ballpark, assembled with totally new technology.

## What is ‘Quantum’?

Quantum mechanics. It’s frequently used by the sci-fi show’s token show-off to demonstrate their knowledge of physics. But what are quantum mechanics, really? As a concept, they’re not that tough to grasp, and you’ve probably witnessed some of the principles in action without even realizing it! For example, have you ever played the game of hiding a coin under one cup, and then shuffling it with two other cups?

Assume someone sits down to pick a cup, and they can’t tell where the coin is based on you, or your observation. Until they pick up a cup, the coin *could *be under all three cups. Basically, there’s a 33.33% chance the coin is under the cup they choose. However, once you pick up the two cups you know are empty, the odds condense. There’s now 100% certainty the coin is beneath the final cup, and 0% possibility it’s under the other two cups.

In real physics, this example doesn’t work perfectly. Most quantum mechanics, once observed, break down into observable truths, and you’re an observer too. You, the shuffler, have some way of knowing which cup the coin is under. The coin is probably making a sound as it’s dragged around the table, or maybe the coin is so heavy it is obvious which cup is holding it. If you know where the coin might be *at all, *it means that there is one observable outcome where the coin’s underneath the noisy cup, and not three potential outcomes where the coin is under all the cups. Observing this makes it true for your opponent, as well!

Assuming coins are actually **particles**, and the cups are really **probable locations**, you’ve got something that gets close to real quantum mechanics in action!

## Make Waves

Quantum mechanics (without *any *math in the explanation) are just a way to explain the probability of a particle existing somewhere in a real, physical environment when it’s actual location can only be expressed *through *that probability, or else it stops behaving the way it’s ‘supposed to’.

This probability breaks down into wave forms, where certain spots are more likely than others to have a particular particle than others. For example, the cups all have a 33.3% chance of coin, but the table outside the cup has a 0% chance of coin. In a dark room, where nobody can observe that the surroundings are coinless, but everyone ‘knows’ coins go under cups, (like we ‘know’ where electrons tend to be found in an electron shell), the chance of it being on the floor are very, very small – but not 0%.

Out of the places you’d pick a coin to be, though, it’s *probably *still under one of the cups, and almost certainly still on the table. If you looked at this probability on a chart, you’d see hills of likelihood where the cups are, and dips where they aren’t! In this way, we calculate the probable locations of things like electrons and photons, which behave in ways humans don’t fully understand yet. The coins in the above example are like those particles! A photon is *probably *in a certain area given what we know about its behavior – but attempting to *actually *measure it as a wave makes it behave like a particle, breaking it’s quantum state. Information is lost, and the particle no longer behaves like it did when it wasn’t being observed. Picking up the ‘cup’ to observe fundamentally changes the behavior of the ‘coin’ underneath!

How does this turn into a revolutionary computing method?

## Entangled

Quantum entanglement describes items (like particles) being tied to each other in such a way that one item can’t be described without also describing the other items in the system, which causes it to collapse as though you were looking at all of it. For example, say you put *two *different coins under *two *cups. Each cup has a coin, but which cup has *which *coin can’t be accurately described until one cup is lifted.

Once that cup is lifted, the first coin is described. The second coin has now also been described because there’s no way the coin you’re *looking at *is under the other cup, and each cup* now *contains/has only contained its respective coin. But only once you observed it. The probabilistic wave forms have now collapsed into two points with 100% likelihood.

That *doesn’t *mean that one coin/particle was always, 100%, underneath its specific cup – until you picked up the cup, both were underneath both cups, mathematically speaking (remember, this is a rough example – coins and particles have different laws attached). Entanglement also has a lot to do with superposition, since both coins would have had to share a location for the cup/coin thing to happen.

## Superposition

Superposition describes things existing in the same space – and it’s not solely tied to quantum mechanics. Two notes played on an instrument at the same time, for example, create a new note out of their superposition. The big thing about superposition is *waves. *Physical objects can’t be superimposed upon one another, and two particles can’t be in exactly the same location. However… properties of objects can be expressed mathematically, in wave forms, and in *that* way they can be superimposed. Much like different wavelengths of light can combine to form a new color, the odds of objects being in a certain state, or being in a certain, unobservable spot can combine in superposition!

In the two-cup example, the coins are in a state of superposition until the cup is removed and their options are solidified; before the cups are removed, whatever equations are used to describe a coin’s location can be added to the equation to describe the other coin, and both equations are still valid. Neither is disproven by the existence of the other until one is observed. *Until *one is observed, the superposition stands.

These concepts, when put together, allow computers to read bits that aren’t yet bits, but *could be *bits.

## Sum Total

All of this sounds really complicated – and it is, mathematically – but conceptually, it just boils down to ‘things can be predicted to be in multiple spots at once’, and ‘things can be a combination of the probabilities of other things, instead of just one thing, until observed’.

A quantum computer looks at probabilistic bits like we look at those coins, and it doesn’t think ‘that’s a 1’ – it thinks ‘this is probably a 1, but if it was a 0, how does that change the data?’ and ‘how does this being a 1 affect later bits?’ The most common path of quantum computing research uses qubits, which stay in a state of superposition.

This means that the qubit is both a zero and a one until the computer looks at it and determines its state via some randomized metric that maintains the quantum state. It could be the state of the electrons at the time the computer reads it, it could be the magnetic direction the qubit is excited into randomly, etc. it just has to behave in a way that outside observers can’t definitively say leads to one specific outcome. If it can manage that, then it can calculate all the available options all at once.

## Advancements

How is this faster, you may ask? Well, the qubit is ‘stacked’ onto other bits. The qubit can be two states, and subsequent qubits can be two states, and… they daisy-chain together to form exponentially larger potential states, which then lead to answers being calculated simultaneously, instead of linearly. In a perfect system, faults are discarded, and then the quantum computer spits out the right answers in a fraction of the time it would have taken a classical computer.

For example, let’s say a password is tied directly to the state of a pair of dice in an automatic shaker. A quantum computer will be able to spit out a probabilistic password, but a classical computer won’t be able to compete! Even if it’s a supercomputer, it will have to get lucky if it wants to guess what the shaker’s results are going to be before the dice are shaken again.

While this sounds very futuristic, websites are already using algorithms to convert random footage into protection for their servers: the lava lamp wall used by Cloudflare is one such example. By the time a classical computer has calculated what the algorithm required when lava lamps A-Z were in *any *position, literally all of them have changed. As a result, the code has changed as well, rendering that math useless. A quantum computer will be able to step up to the plate where the classical computer has struggled!

As Dr. Shohini Ghose puts it, this isn’t the equivalent of several classical computers, or one big classical computer compressed into a smaller state – it’s a totally new technology that will behave differently as it advances. Even a super computer would struggle with the lava lamp wall! However, quantum computers may not. Every qubit used to calculate has the potential to lead to a correct answer, or a wrong one. Good quantum computing will kick out incorrect answers as soon as they’re produced, and you’re left with something that the lava-lamp wall algorithm will take as an answer.

Dr. Ghose uses the example of a coin-flip game, where participants face off against a quantum computer. If the computer is told to win, and it goes first, it produces a probabilistic result that only collapses with the other player’s input – the computer is essentially allowing its coin to continue spinning in the air until it can tell what the human player has, and then it catches it, to spit out the answer that it *always *had. The answer existed in a probabilistic state – and it won, it just needed to be observed to tell the human that. The computer only *loses *when it mistakes the ‘noise’ answer for the actual result. If it were able to successfully suppress noise, it would win 100% of the time.

## Why Not Earlier?

These computers have been seriously considered as a project since the 80s and 90s. And now, they’re making a resurgence. What kept them from being considered earlier?

Logical faults are a big part. Modern AI can suppress things it knows aren’t ‘really’ part of an equation’s answer, but the coin-flip computer above still lost 7% of the time to bad answer output. In the past, quantum computers wouldn’t have been able to correctly identify their own mistakes even down to 7% without a classical computer running alongside them, which defeats the purpose. Unlike classical computers, where faults like that come from the hardware, quantum computers are getting these errors from the state of universe itself. Of *course *that’s difficult to compensate for.

Aside from that, there were also mechanical issues to sort out first. The computer can’t be allowed to turn the qubit into a regular bit, which is called ‘decoherence’. Decoherence happens once the system is connected to something measurable, observable: out of two cups, lifting one solidifies the probability, and the other cup, even though it hasn’t been observed, definitely has the other coin. If it’s solidified into a regular bit, it may as well have not been a qubit at all!

Mechanically, to avoid decoherence, speed and environmental controls are *essential. *In quantum computing, you aren’t maintaining that quantum state indefinitely – the longer the computer has to maintain that, the worse off the state is, until eventually something collapses in a measurable way. Heat will do it, stray magnetic or electricity pulses will do it – flip one qubit, screw up the system or collapse it entirely. Decoherence has destroyed the calculations.

Side note: if you’ve heard of the double slit experiment, that’s an example of decoherence! Measuring the particles breaks the system while deliberately** not** measuring them allows for that nice waveform. Their final location becomes known, but not the path they took to get there. In computing, measuring the qubit before the computer gets to then breaks it down into a not-qubit. Rendering the system decoherent, and screwing up the results of the calculations.

## Tid-Bit

Ironically, Schrodinger *haaated *that his ‘cat experiment’ got big because folks were taking it too literally. For those of you who haven’t heard of the thought experiment (no cats were ever actually put in a box) the experiment’s set-up was that radioactive material has a certain % chance every second to release a radioactive particle, and then putting this material next to a particle-sensitive trigger would release poison via that trigger into the cat’s box. If there’s no guarantee of poison being released into the box, there’s no mathematical certainty that the cat’s either alive or dead, so it’s *both.* Just like the coin is under all three cups.

* *But not really. At the scale the experiment would have to take place, the cat’s as good as already poisoned (a lump of radiation has so many individual atoms that the odds of one *not *releasing a particle at any one moment is basically zero), but Schrodinger was struggling to explain the concept to laypersons who otherwise had no exposure to physics.

The thought experiment does a great job of breaking down what’s *actually *occurring with superposition. It’s not about the cat, or poison, it’s about the particles. If the experiment could be particle-sized, it *would *work the way it’s described.

Sources:

https://indianapublicmedia.org/amomentofscience/the-heisenberg-uncertainty-principle.php

https://www.sciencealert.com/quantum-computers

https://jqi.umd.edu/glossary/quantum-superposition

Shohini Ghose via TED Talk (direct link: https://www.youtube.com/watch?v=QuR969uMICM)

https://www.ibm.com/quantum-computing/learn/what-is-quantum-computing/

https://www.nature.com/articles/s41598-020-75730-1