From what I understand, they will be able to calculate difficult equations FAR faster than current computers. Cool. But what is this actually useful for? I saw some scientist proclaim that quantum computing would solve food issues and lead to cancer cures.
How??
Comments
Quantum computers can theoretically break much of modern cryptography. That’s basically it, the usefulness of everything else is speculation.
Your first statement is completely incorrect. They have the potential to solve one family of problems much faster than regular computers but it hasn’t been done yet and the only obvious practical use right now is cryptography.
And I don’t know who this “scientist” you mention is but it sounds like bs.
State of the art modern encryption is theoretically breakable, the problem is that in practice it would take millions to billions of years to crack an encrypted message with current methods. Quantum computing has the potential to reduce that down to a manageable time frame, which in turn would mean that encryption was no longer reliable
A traditional (“classical”) computer has a number of problems that could technically be solved using them, but the energy required to do so would boil the oceans, not to mention the amount of time that it will take (longer than the time currently elapsed since the big bang).
A quantum computer is a bit different; there are a few classes, however one of the simplest to visualise is “quantum annealing” where the machine is configured with inputs and allowed to “settle” into a low energy state, and that state is the solution to the problem. You can repeat this a few times and come to a conclusion without needing to perform all of this intermediate steps, it’s just a result of the way the system operates (See D-Wave).
Recently 3 Blue 1 Brown did an intro to quantum computing video that does a decent job of explaining the what and how using an actual example, but its definitely above an eli5. Worth checking out if you want more of an eliInterestedInMathHighSchoolerOrAbove
They can do math in entirely different ways than normal computers. This allows for some types of calculations to happen at exponentially faster rates (things that were considered impossible due to time constraints suddenly become simple). I’m sure there are a lot of very smart people putting together the formulas they want to run on quantum computers, but what we will actually be able to do with them once they are cost effective is yet to be seen. For now, people have been able to use creative methods to solve some problems with normal computers faster than existing quantum computers can manage, but that is expected to change as quantum computers improve and become affordable. And eventually people will come up with creative methods for quantum computers as well.
One thing we are already certain of is that solving the math used for many existing methods of encryption becomes trivial on a quantum computer. Many old, encrypted files and databases will suddenly be available to the world once quantum computers are affordable. Anything that has ever been leaked that was encrypted with most standard methods will be easily decrypted. There are certainly people holding onto piles of old data dumps, waiting for the day they can access them. Companies are slowly moving to encryption methods that we have not yet figured out how to quickly crack with quantum computers, but anyone who hasn’t completed that by the time quantum computing is widely available is going to suffer. Also, due to the existence of old data dumps, anyone who hasn’t changed their passwords in a while will be at risk regardless of updated encryption.
Hopes about solving real life issues in the world come down to predictions about what we will be able to model with the computers. They will very likely become standard for modeling molecular interactions of things involving both genetics and medicine, potentially leading to foods that grow more easily and cheaply and cures for all kinds of diseases.
Claims about quantum computers are waaay over blown. The very far fetched idea that scientist was referring to is that quantum computers could possibly be used to simulate molecules therefore leading to the discovery of new drugs, new materials that would revolutionize technology.
At the same time, we don’t know what the future holds for them and what they will actually do. No one predicted Minecraft when they wrote the first boolean algebra papers
That’s a great question, little buddy! Grab a juice box and let’s dive into it!
You know how in your activity books there are mazes where you draw a line with a pen to find the way through? Well, some really smart people have built robots that are really fast at drawing that line and finding a way through. We call them computers but we’ll call our little robot Chip. Hi, Chip!
But hold on! Now let’s imagine the maze is really really big. Like REALLY BIG! Bigger than the entire planet! Well, even though Chip is really really fast, this huge maze is too big for him! It would take him waaaaay too long to find a way through.
But Chip has a secret magic friend, Wally. Wally is a quantum computer, kind of like a strange being that’s made of water. You know how if you tip over your water glass, the water goes everywhere? Well, that’s Wally.
So, to find a way through the huuuge maze, Wally doesn’t have to go back and forth like Chip, trying to test different paths. Instead, Wally can pour himself into the maze like tipping over a glass of water, and he can flow everywhere at once! He’ll travel like the waves in the ocean at the beach, testing every path at the same time. The waves that end up in a dead end, they’ll cancel each other out, and the only wave that’s left, is the path that leads out of the maze! Now Wally can help Chip to the other side! Thanks, Wally!
Ok, that’s enough quantum physics for now, little buddy, it’s past your bedtime already. Let’s get you to bed!
There’s different classes of difficulty, regarding how hard a math problem (more accurately here, algorithm, is to solve). Some are easy, like checking if a sudoku game is solved: just check if the numbers don’t repeat. Those are polynomial problems. Some are really hard, like actually solving sudoku: that’s a non polynomial problem.
Modern traditional computers cannot solve non-polynomial problems. Or, well, they can, but it takes a LOT of time and energy (a non polynomial amount, actually).
Quantum computers work differently. They’re no faster than traditional computers for the easy (polynomial) problems, but they’re capable of solving certain hard (non-polynomial) problems in a reasonable time.
Besides sudoku, some noteworthy non polynomial problems are protein folding (which would impact drug production for example), and the prime factoring problem (which would break cryptography).
They’re not faster computers, they’re computers that solve different problems, problems for which we currently have no reasonable solutions
Quantum computers are (potentially) faster at certain problems. They can:
Factorise numbers exponentially faster than classical computers. This is important in security, because most of web cyber security is based on the RSA algorithm which assumes factorising numbers is hard.
Can black box search much faster than a classical computers. This is important in, say, Bitcoin mining where you could mine a Bitcoin considerably faster than on a classical computer (up to 2^128 times faster in some scenarios). Could also have security implications (but these are more easily solved).
Can create untappable security channels. Two quantum computers can create a link between them that cannot be read by someone else without both original computers knowing. This effectively prevents man in the middle attacks.
Can create smaller transistors, in fact quantum computers support single atom transistors, which classical computers could not support due to quantum effects getting in the way.
Can simulate quantum systems exponentially faster than classical computers, which is more important in research than commercial use
A bunch of other proposed ideas, including faster graphics processing and faster AI processing
So far the only example of “quantum supremacy” (where a quantum computer does something better then a classical computer) currently (that I know) is Googles quantum computer which could detect whether white nose was true random or psuedo-random in a few minutes which would take a classical computer, following the same algorithm, something like a billion years.
It’s not about them being faster. It’s about them being able to do more. It’s like humans digging single shovel sized holes – two people would dig the same size of hole together as if one person dug it, maybe even smaller since the second person interrupted the first, but together they dug two holes instead of one.
Quantum computers are like adding extremely huge number of people at the task. Doing a lot in parallel, but it doesn’t mean it will accelerate a single operation. They would help with such tasks like testing all possible passwords all at once. Testing one password is short, but there are a lot of them to test, it would take nearly eternity to do it one by one. But good enough quantum computers could try all of them at once. That would totally break any encryption and our network infrastructure relies on that. Additionally having means to do such brute force approach would mean we could do things like more precise physics simulations that simulate individual particles. We could do more “small tasks at once” which would open a lot of doors.
They can simulate complexity very well. So well that it makes today’s processors look like a steam-cranked addition engine from the 1800s.
When you can simulate things with quantum power, you can include enough variables and calculate fast enough that you can answer questions that are impossible otherwise.
Cancer cures? Let’s simulate chemical actions on the sequenced DNA and include every possible alternative, then go 1,000,000 steps beyond what’s possible today in terms of mapping the chemical and systemic interactions.
Instead of taking a century, you could do this in a few hours.
Quantum with AI to actually translate our questions into algorithms is a further multiplier.
Quantum computers aren’t like normal computers, and can’t simply do the same things but faster. They’re entirely different, but can solve some problems far faster, and there’s one type of problem where this can really change things, and that’s called an NP problem.
NP problems can only be solved by guess and check. Like brute forcing a password. You just have to look at every possible solution, and check to see if it’s correct. Quantum computers operate like schrodingers cat (which is both dead and alive until you observe it, though an oversimplification). A function is a quantum computer doesn’t really give one solution, but a probability distribution of all possible solutions. As the cat is both dead and alive, the quantum output is a superposition of all possible answers, but it turns out, the correct answer will have just a tiny bit higher probability than all the others, and there are operations you can do on the probability distribution to widen the gap between the most probable solution and all the rest. So repeatedly performing these operations will change this very slightly greater probability into nearly 100%.
The deal is that certain computations which take exponential time with classical computers may take linear time with quantum computers.
That means that certain stuff that it’s possible to compute but would have taken too much time to be computed, can be computed in a more practically reasonable time.
The most common example is breaking classical cryptography algorithms, which were based exactly on the fact that breaking them needed more time than any attacker would have.
There are many applications however, in simulation/modeling, real time control, scalability.. basically anything which is possible to compute in principle, but whose infeasibilty is currently due to the exponentially growing time that it would take to compute it.
It took 13 years to sequence the human genome. A quantum computer would only take a minute. Thats fast but it doesnt take into count all the students and researchers who dedicated parts of their life to this research. The building they had to work in. The collaboration with other departments. Gallons of coffee.
they’re very good at problems where it’s hard to find the solution, but easy to check which solution is right. They’re also WAY overhyped in a big way. Odds are no normal day to day use will ever come from quantum computers, but they do allow a lot of interesting new algorithms and novel solutions to problems that we couldn’t previously brute force
its in development and they have no real world use right now, so only research
A utility-scale quantum computer could simulate some complex chemical reactions in significantly less time than a “classical” super computer (this has already been done with small carbon molecules), leading to the development of, for example, new pharmaceuticals and fertilisers.
I’m going to go really basic with it.
Let’s say I give you an equation like y=35x^1.95890 and tell you to find all values of x for y being prime whole numbers through a trillion.
You (or a standard computer) will go through these millions of calculations one-by-one until you’ve found each solution.
A quantum computer with enough qubits will simultaneously solve all of them in the same few compute steps. And actually, it’s kind of calculating for all numbers and just returning correct solutions to your parameters.
VAST oversimplification, but it can be carried into real-world solutions really quickly.
Cryptography is the big one. If you need the correct private key to break encryption, then you can either type in trillions of combinations of passwords, or you can just calculate every possible string at the same moment and see which one produces a matching hash through the encryption algorithm.
Medicine will likely be changed by quantum computing of protein folding. If a medication is needed that bonds to a certain receptor on a certain cell, and we need to artificially produce a protein that’s the correct folded shape, then calculating in reverse from every folded shape that fits to protein strings at once is much faster than individually testing proteins and hoping we find one that bonds correctly.
It is like this, though not exactly this:
Traditional CPU can only solve “yes” and “no,” so if they’re solving an equation that has dependent variables, it has to solve them one at a time, and won’t know if it got the first one right until it gets to the end. Then it has to start over from the beginning if it got any variables wrong.
Quantum computers can solve with a third option: “maybe?” and this allows it to solve the variables in the equation simultaneously, because even though they’re dependent, the quantum computer doesn’t have to start from the beginning every time.
Modeling quantum mechanics currently is hard. Even the best super computers can’t do it perfectly. There’s just too many possibilities so you make simplification.
Quantum computers will be able to model quantum mechanics a lot more easily. It could speed up development of drugs and new materials because you now have a computer that can model molecules, which is what drugs and stuff are made of.
Why is modeling stuff important? Because that’s the way we put our understanding to use and models help us do things. For example, predicting the weather is modeling. The weather forecast can be wrong because models aren’t perfect. Now imagine if the model were perfect.
Is this a Seinfeld bit?
The ELI20 version of how quantum computers work is that quantum computers can solve problems that scale in complexity according to N! in a constant amount of time, as long as they have enough (some factor of N) qbits. Those same problems running on a classical computer would require an N! amount of time, so they quickly become too large to solve exactly no matter how fast your classical computer is.
So the real question is, what sort of important problems scale like this in the real world?
Encryption – explained at least as well as I can elsewhere in this thread.
Computational Chemistry – The exact solution to the schrodinger equation which describes the energy of a molecule based on it’s chemical structure and can be used to predict useful properties about a substance, scales as N! where N is the number of electrons in the system. This could improve computationally aided design of new products in multiple fields if the classical approximations are too slow or too inaccurate to be useful (energy, medicine, manufacturing, etc.).
Logistics – The “Traveling salesman problem” is a classical example of a N! problem. Given a list of cities, can you find the shortest route that allows the salesman to stop at each city exactly once? Similar sorts of optimization problems crop up all the time in logistics, and quantum computers could solve larger versions of them much quicker.
The real challenge is finding problems that don’t scale as N! that can still benefit from quantum computers. That makes it harder to predict the role quantum computers will have in every day computing decades from now. They could be niche instruments that sit in university, corporate and/or national labs that are only used for specific problem types, or it could be something that gets commoditized to the point where you’ll have a device you plug in at home and connect to with your phone/laptop/etc when quantum calculations are needed.
Everyone is entirely missing the point of the question.
The issue with answering it is that it takes a high-level expert in a field. I don’t know what the bottlenecks of chemistry are but a chemist with 20years experience might instantly think of an answer.
But one example actually is chemistry. There are countless of combinations of molecules with varying negative/positive charges. Each behaves uniquely. A quantum computer can test all of these for a specific result at once instead of the chemist having to grind out tests for years
> From what I understand, they will be able to calculate difficult equations FAR faster than current computers. Cool.
They can solve some problems exponentially faster.
> But what is this actually useful for? I saw some scientist proclaim that quantum computing would solve food issues and lead to cancer cures. How??
The biggest use case for quantum computers is simulating quantum mechanics. That means you can accurately simulate atoms. Imagine being able to simulate the interaction of drugs in the human body. You could test thousands of variations of your drug before you need to do actual physical tests.
a powerful enough quantum computer could efficiently model things like formation of crystal structures (useful in material science) or find certain proteins which fold in particular ways (idk im not a biologist) that would make them useful for certain treatments or to further study things like prion diseases, another potential use (which i personally find cooler) is that they could perform shors algorithm in polynomial time which is a fancy way of saying they could break RSA and other widely used encryption schemes reasonably quickly which is quite dangerous for cybersecurity if encryption standards aren’t shifted to post quantum ones (ie stuff that quantum computers can’t break as easily) before someone with bad intentions makes a sufficiently strong quantum computer
I’ll have a go: when you count pokemon cards, you want to know how many you’ve got. If you wanted a computer to tell you, it would take a long time to ask it the question, because the computer doesn’t know anything about pokemon cards except what you’ve told it, so:
You need to tell it what a pokemon card is in enough detail that it doesn’t count nba cards as pokemon. And you need to ask it your question.
So computers are already good at counting, they’re so good that they can count massive numbers of lots of different things and store those things they’ve counted in something called a database. But like you and me, they can’t keep all of that in their head at once, and get tired when they try.
A quantum computer can keep more, a lot lot more in its head at once, and its really fast. So fast that it can work out a really difficult secret code straight away that it would take a normal really big computer thousands and thousands of years to work out.
Because it can hold more in its head at once, and because its fast, we can ask it to think about lots and lots of really massive databases, it could possibly even work out how to cure cancer, and run test scenarios, from really random data really quickly.
At the moment we have to be careful about the data we ask computers to think about because otherwise it would take too long. With quantum computers, we don’t need to be so careful, and we can ask the computer to answer questions we might not have thought about.
My tone felt really patronising, but you are 5, and I’m an adult and I can make stuff up if I like. 🙂
From what I’ve looked into as a computer science guy, right now we compute by comparing memory locations and the patterns of bits stored there. So to do a calculation that requires trillions of manipulations you need to manipulate many memory locations, trillions of times.
With quantum you create a single memory location with enough qbits to answer your question (this is the hard part) then you program how this location should be changed, then you add stuff to it. So by coding the process of manipulation to emulate your problem, you look at the final state of the qbits to find a value that would have taken too long with traditional compute to figure out
You are underestimating what “far faster” means. There exist some problems we theoretically know how to calculate, but there is a little issue that it would take trillions of years to actually calculate them on a classical computer. Some of those problems a quantum computer should be able to solve in a reasonable time frame.
If effect, quantum computers would be able to solve some currently impossible problems, it would push the boundaries of what can be computed.
And that’s a big thing, similar to how AI is currently a big thing because it’s solving computational problems that were previously impossible.
The big deal with quantum computers is that they generate real quantum states, from which information can be extracted. Classical computer systems can only approximate these.
The actual information gained from measurement is usually small (in comparison to the whole state space), but unique. By increasing the number of qubits, the complexity of the quantum state increases which means the potential for more information to be extracted. A small number of qubits can store complexity that requires a extremely large number of traditional bits.
The generation of complex quantum states and the information researchers may extract then, hopefully, leads to new/better understanding of quantum mechanics. Some real examples are calculating energies of small molecules and simulating magnetic properties of atoms. Information gleaned may then have real-world applications such as reducing time needed to synthesize a new chemical, simulating drug interactions, finding alternative applications for compounds, etc. This could then have impact in agriculture, pharmaceuticals, logistics, battery technology, and more.
A misconception of quantum computers is that it gives you “the” answer to something automatically. Researchers must design a specific measurement to get the information they want. Oversimplifying, quantum computers generate a “probability map” of answers and by carefully measuring the output (then usually repeating many times) the most probable answer is found. Quantum computing still relies on classical computation and/or experimentation to verify answers are actually correct.
Essentially, if you want to learn more about how the “quantum world” works, it’s advantageous to model and learn from real quantum states instead of relying in simulated states.
as other comments have said quantum computers could (in theory at least) do massively complex calculations much faster than current top-end computers.
think about super computers filling whole rooms, and then a quantum computer could be no bigger than your laptop (im exaggerating – we dont have any yet but, this is the kind of change in scale we could see)
why is this actually useful?
well, a lot of your personal data, banking info, on line accounts..etc.. is made as safe as possible using cryptography. this is basically using complex maths to hide your info. Maths that current computers could take decades to crack/hack into
a quantum computer renders all that obsolete -it could hack anything in seconds
so …banks are shitting themselves
Let’s say our computers are like cars. A quantum Computer wouldn’t be a super car. It would be a boat. Opening up new possibilities that would’ve been impossible before.
They don’t do most computation faster than conventional competition, but they do one algorithm very very very well, which is factoring primes. Nearly all computational security until recently has involved the multiples of 2 very large primes, because it’s difficult to do normally.
Every computer that uses this becomes vulnerable with quantum computing. All of your passwords will be vulnerable. All bank passwords will be vulnerable. Hopefully they have already upgraded to quantum proof security.
It’s probably going to be like Y2K, not many important systems will go down because of quantum computing, but it’s possible that it could also be much worse.