Has The Age Of Quantum Computing Finally Arrived? - General

[video=youtube;KioLI5M4iBc]https://www.youtube.com/watch?v=KioLI5M4iBc[/video] Read about the prototype quantum computer created by IBM that you can use [b]RIGHT NOW[/b] over the internet [url=https://www.research.ibm.com/ibm-q/]here[/url]. Click the 'experiment' section to get links to the software and documentation [b]Play the [url=https://www.research.ibm.com/ibm-q/quantum-card-test/]interactive card sorting demo[/url] to see how quantum computing works[/b] Quantum computers are special types of computer processors that run on [b]'qubits' (or 'quantum bits')[/b] which employ the funky physics of quantum mechanics to perform certain types of calculations much more quickly than traditional computers. [b]IBM has made a prototype using 5 qubits which they have connected to the internet so that anyone can try it out online.[/b] So far, this toy prototype has been used for experiments such as simulating two hydrogen atoms (which cannot be done easily on a traditional computer due to the quantum nature of the electron orbits). Quantum computers require specific laboratory conditions, including an extreme temperature controlled environment which is colder than the depths of outer space. So, [b]It is unlikely that your phone will ever be a quantum computer in the foreseeable future[/b]. However, quantum computers can be [b]linked to any traditional computing device over the internet via cloud technology[/b], meaning any device could functionally become a quantum computer. According to IBM, simple quantum computers could have the following specific uses in the near future: [QUOTE][b]Medicine[/b] Untangling the complexity of molecular and chemical interactions, leading researchers to the discovery of new medicines and materials.[/QUOTE] [QUOTE][b]Supply Chain[/b] Finding the optimal path across global supply chains to deliver a product to your door with the most effective balance between fuel and time. [/QUOTE] [QUOTE][b]Financial Services[/b] Finding new ways to help model financial data and isolate global risk factors to make more informed investment decisions. [/QUOTE]

the card sorting demo would be a lot more impressive if there were more than 4 cards, but they probably don't want to overwork their system sorting cards [editline]28th October 2017[/editline] our phones won't directly be using quantum computing but you can be sure there will be things you can access from your phone that will be quantum computed

As with any rhetorical internet question video, the answer is "sort of, but of really" Quantum computing is fucking amazing, don't get me wrong. [B]But, the IBM demo doesn't actually use real qubits, it just simulates what it would do. [/B]EDIT: See below for software to interact with the real thing.

I think development of mainstream quantum computers will begin now, but we won't see it really in the mainstream for quite some time. I saw a few videos of actual quantum computer needing to be kept in below freezing point temperatures to operate, which tells me that we won't see them in every people's homes, but we can definitely see them in bigger businesses and eventually for cloud services (like AWS).

[QUOTE=Gbps;52832455]As with any rhetorical internet question video, the answer is "sort of, but of really" Quantum computing is fucking amazing, don't get me wrong. But, the IBM demo doesn't actually use real qbits, it just simulates what it would do... Which shouldn't be surprising, but the post makes it out like it's really using quantum computing to calculate the result.[/QUOTE] You CAN use the actual quantum computer if you download the software: [url]https://quantumexperience.ng.bluemix.net/qx/experience[/url] You need to [url=https://github.com/QISKit/ibmqx-backend-information/blob/master/backends/ibmqx4/README.md]understand the documentation[/url] and use python though.

[QUOTE=Zyler;52832461]You CAN use the actual quantum computer if you download the software: [url]https://quantumexperience.ng.bluemix.net/qx/experience[/url][/QUOTE] Neat, noted. You should clarify that in your post!

I can't find the paper right now but someone actually managed to get a very simple renderer (it integrates to a black and white gradient iirc) to work on their quantum computer. Their presentation was insane. [editline]29th October 2017[/editline] found it [media]https://vimeo.com/180284417[/media]

The demo didn't do anything to explain how quantum computing works. "Binary computer has to go through the cards until it finds the correct one, but quantum computer finds it on first try. There you go."

Friend at university is heavy into this, has been attending conferences all over the world and has been literally pestering PhD students and groups to get into research. I'd still say quantum computing has got a signficiant distance to go,but the groundwork is definitely there. There's just a few obstacles we've to overcome, such as protecting the decoherence of quantum information during processing and more efficient and effective ways to store and manipulate qubits - the current methods (eg. laser traps ) are pretty finnicky and require a stupid amount of cooling; something that's a huge block to getting in a wider area. We aren't going to be seeing personal quantum computers or quantum computers being bought by companies, it seems to me that quantum computers will remain entirely in the cloud; companies and users will pass on data and the relevant quantum information to a quantum computer in a lab somewhere to be processed. Quantum encryption and all that jazz will ensure a stupid amount of security.

I still don't get how the fuck the quantum computer gets the card trick in one attempt. Like my brain can't wrap around it? It has zero information to work with, other than the fact there are 3 Kings and 1 Queen. It'd be plausible to my brain if it at least had to turn one card over, but it can do it without turning any over? Seems like witchcraft, can someone do a better job of explaining?

[QUOTE=loopoo;52833141]I still don't get how the fuck the quantum computer gets the card trick in one attempt. Like my brain can't wrap around it? It has zero information to work with, other than the fact there are 3 Kings and 1 Queen. It'd be plausible to my brain if it at least had to turn one card over, but it can do it without turning any over? Seems like witchcraft, can someone do a better job of explaining?[/QUOTE] Quantum computers are massively parallel probabilistic machine. Imagine the quantum computer calculates the probability of each of the cards being the queen in parallel, in a single operation. The quantum computer then returns a set of probabilities for each possible cards being the queen. You can then pass this resulting set to a classical computer, and using it to select the card with highest probability. If the quantum algorithm was designed correctly, the card with the highest probability should be the one being a queen.

[QUOTE=B!N4RY;52833201]Quantum computers are massively parallel probabilistic machine. Imagine the quantum computer calculates the probability of each of the cards being the queen in parallel, in a single operation. The quantum computer then returns a set of probabilities for each possible cards being the queen. You can then pass this resulting set to a classical computer, and using it to select the card with highest probability. If the quantum algorithm was designed correctly, the card with the highest probability should be the one being a queen.[/QUOTE] What difference does that make? The probability is the same 1 out of 4 regardless of how you calculate it.

[QUOTE=Talishmar;52833208]What difference does that make? The probability is the same 1 out of 4 regardless of how you calculate it.[/QUOTE] The output is a probabilistic result for a quantum search algorithm. Rather than generating a result of "Card 3 is the Queen" like classical computers would, a quantum computer would generate a result something along the lines of: "Card 1 has a probability of 1.65% of being the queen; card 2 has a probability of 2.18% of being the queen; card 3 has a probability of 96.17% of being the queen; card 1 has a probability of 1.17% of being the queen" Quantum computers work substantially different than classical computers, and its programming models are completely unrelatable by classical computers. Currently, quantum algorithms are modelled using [URL="https://en.wikipedia.org/wiki/Quantum_circuit"]quantum circuits[/URL], built by sequential transformations using various [URL="https://en.wikipedia.org/wiki/Quantum_gate"]quantum gates[/URL]. In a nutshell, you initialize a set of qubits with some initial values, you then put them through a series of transformations, and finally you measure the qubits. The measured qubits at the end will yield a set of probabilities of each qubit being relateable to the answer in some way.

[QUOTE=B!N4RY;52833236]The output is a probabilistic result for a quantum search algorithm. Rather than generating a result of "Card 3 is the Queen" like classical computers would, a quantum computer would generate a result something along the lines of: "Card 1 has a probability of 1.65% of being the queen; card 2 has a probability of 2.18% of being the queen; card 3 has a probability of 96.17% of being the queen; card 1 has a probability of 1.17% of being the queen" Quantum computers work substantially different than classical computers, and its programming models are completely unrelatable by classical computers. Currently, quantum algorithms are modelled using [URL="https://en.wikipedia.org/wiki/Quantum_circuit"]quantum circuits[/URL], built by sequential transformations using various [URL="https://en.wikipedia.org/wiki/Quantum_gate"]quantum gates[/URL]. In a nutshell, you initialize a set of qubits with some initial values, you then put them through a series of transformations, and finally you measure the qubits. The measured qubits at the end will yield a set of probabilities of each qubit being relateable to the answer in some way.[/QUOTE] How can the cards have differing probabilities? Am I repeating myself?

[QUOTE=Talishmar;52833260]How can the cards have differing probabilities? Am I repeating myself?[/QUOTE] mate I feel you, I think we're just too peasant to understand this futuristic technology quantum mechanics is fuckin whack, yo

[QUOTE=Talishmar;52833260]How can the cards have differing probabilities? Am I repeating myself?[/QUOTE] Am [B]I[/B] repeating myself? Did you even read through my post? Like I said before, I'm talking about the output of a quantum search algorithm for finding the card. You can also interpret the output as a set of [U]confidence [/U]that the computer believes in each card being the queen [U]after [/U]running the search algorithm. 1/4 is the chance of a card being the right one, if you were to pick one at random. It has zero relevance at all here. In classical computers, you can iterate through the deck and check if each card is the queen. You would get an output of : Card 1: Not a queen Card 2: Not a queen Card 3: A queen Card 4: Not a queen Except for quantum computers, the answer isn't always a straight forward YES or NO (or a digital 1 or 0). It's expressed as a complex number, but we'll pretend it's a real value ranged from 0.0 to 1.0 for simplicity. The quantum computer searches all 4 cards in parallel, and returns a list of numbers highlighting its confidence in believing which card is the queen, after being searched by the quantum search algorithm. You can interpret an output like this: Card 1: I'm 1.65% confident this is the queen Card 2: I'm 2.18% confident this is the queen Card 3: I'm 96.17% confident this is the queen Card 4: I'm 1.17% confident this is the queen You can therefore conclude card 3 is the right answer.

[QUOTE=B!N4RY;52833304]Am [B]I[/B] repeating myself? Did you even read through my post? Like I said before, I'm talking about the output of a quantum search algorithm for finding the card. You can also interpret this as a set of confidence that the computer believes in each card being the queen after running the algorithm. 1/4 is the chance of a card being the right one, if you were to pick one at random. It has zero relevance at all here. In classical computers, you can iterate through the deck and check if each card is the queen. You would get an output of : Card 1: Not a queen Card 2: Not a queen Card 3: A queen Card 4: Not a queen Except for quantum computers, the answer isn't always a straight forward YES or NO (or a digital 1 or 0). It's expressed as a complex number, but we'll pretend it's a real value ranged from 0.0 to 1.0 for simplicity. The quantum computer searches all 4 cards in parallel, and returns a list of numbers highlighting its confidence in believing which card is the queen, after being searched by the quantum search algorithm. You can interpret an output like this: Card 1: I'm 1.65% confident this is the queen Card 2: I'm 2.18% confident this is the queen Card 3: I'm 96.17% confident this is the queen Card 4: I'm 1.17% confident this is the queen You can therefore conclude card 3 is the right answer.[/QUOTE] So in very basic terms, the quantum computer looks at all the cards at the same time to get a general idea what they look like, and then can point to which most looks like the queen?

That's where I was getting hung up, I thought the entire point of the exercise was that the cards stayed face-down until reveal. I didn't realise it was basically saying: classical computers go through one by one, whereas quantum computers can look at all of them at the same time. It now makes sense.

[QUOTE=Talishmar;52833316]So in very basic terms, the quantum computer looks at all the cards at the same time to get a general idea what they look like, and then can point to which most looks like the queen?[/QUOTE] Close, but there is no "generalization" step. Quantum computer wouldn't be solving the problem like neural networks, if that's what you're thinking towards. A more accurate simplification would be a quantum computer looks at the input, work out every single possible solution in parallel, and outputs a list of probabilities of each possible solution being a right answer. Keep in mind that in some problems there could be multiple correct answers, so quantum computers excels very well in such areas.

[QUOTE=Talishmar;52833260]How can the cards have differing probabilities? Am I repeating myself?[/QUOTE] It doesn't know before it checks, just like the traditional computer. It's the finding algorithm that is different. In a traditional computer, it would check them one at a time until it finds it, but in a quantum computer it would check them all at the same time. The problem is that since this model is based off of particle physics, it isn't a yes/no answer; that is, it has a certain confidence at it being there. This is the hard part to explain. Quantum computers work in such a way that it's far removed from anything like a traditional computer, and they'd probably suck at a lot of problems that traditional computers are great at. [editline]29th October 2017[/editline] [QUOTE=Talishmar;52833316]So in very basic terms, the quantum computer looks at all the cards at the same time to get a general idea what they look like, and then can point to which most looks like the queen?[/QUOTE] Not quite. It does look at all of them but it doesn't see things as there or not there, so there is no 'this looks like a queen,' it's more along the lines of interpreting the existence of the queen itself as a wave function rather than as a solid, physical object, if that makes any sense. As I said, this is really hard to explain. [editline]29th October 2017[/editline] Still a poor example, B!N4RY put it way better.

Quantum mechanics is probabilistic while classical mechanics is deterministic Classical: "Thing is at place x at time t" Quantum: "Thing has a chance of being between a and b at time t"

[QUOTE=NixNax123;52833756]Quantum mechanics is probabilistic while classical mechanics is deterministic Classical: "Thing is at place x at time t" Quantum: "Thing has a chance of being between a and b at time t"[/QUOTE] I don't think anyone is disputing that fact. I think what's getting people hung up is the quantum search algorithm. Namely, how does the computer go from a shot-in-the-dark "Card 1 has 25% chance to be the queen; Card 2 has 25% chance to be the queen; Card 3 has 25% chance to be the queen; Card 4 has 25% chance to be the queen", to the high- confidence answer of "Card 1 has 1.65% chance to be the queen; Card 2 has 2.18% chance to be the queen; Card 3 has 96.17% chance to be the queen; Card 4 has 1.17% chance to be the queen." Which I don't have an answer for. I'm stuck on how the quantum search algorithm works, too. I tried to look it up once. I left very confused and with a mild headache, after reading a university lecture on the search algorithm, which was at the end of a 12-week quantum mechanics course and assumed I already knew all the nomenclature and algorithms taught previously. :v:

will a quantum computer finally be able to help me know the probability of getting a gf?

[QUOTE=loopoo;52833809]will a quantum computer finally be able to help me know the probability of getting a gf?[/QUOTE] 1s and 0s work just fine for that 0 [QUOTE=ForgottenKane;52833556] The problem is that since this model is based off of particle physics, it isn't a yes/no answer; that is, it has a certain confidence at it being there. This is the hard part to explain. Quantum computers work in such a way that it's far removed from anything like a traditional computer, and they'd probably suck at a lot of problems that traditional computers are great at.[/QUOTE] because of that I think that the closest we'll get to having them in normal consumer computers is on a small chip dedicated to certain tasks that just make sense to do on a quantum computer. kind of like how AES encryption is done

[QUOTE=thelurker1234;52834867] because of that I think that the closest we'll get to having them in normal consumer computers is on a small chip dedicated to certain tasks that just make sense to do on a quantum computer. kind of like how AES encryption is done[/QUOTE] Realistically, it's going to be the other way around. Classical consumer computers will coexist on its own, and quantum computers will exist with classical coprocessors built into it. Roughly speaking, quantum computers only excels at very large problems that are massively parallelizable. Such computation power is only applicable in scientific and enterprise settings. For everything else, your classical computers would do much better. That's why many quantum algorithms have both quantum parts and classical parts to it.

[QUOTE=B!N4RY;52835040]Realistically, it's going to be the other way around. Classical consumer computers will coexist on its own, and quantum computers will exist with classical coprocessors built into it. Roughly speaking, quantum computers only excels at very large problems that are massively parallelizable. Such computation power is only applicable in scientific and enterprise settings. For everything else, your classical computers would do much better. That's why many quantum algorithms have both quantum parts and classical parts to it.[/QUOTE] Additionally, while the set of problems that [b]are[/b] massively parallelizable, to the point of being practical candidates for quantum solving, is respectably large; it is, at the same time, a [b]far[/b] smaller set of problems than people often like to think.

best way to describe it is not with cards, that's confusing best way is searching through a list. Say you've a phone directory, and you want to check whether gregg fuckhead's number is in the list. A classical computer would have to iterate through every single entry, eg. "1) I've got george's number. It's not gregg's number 2) I've got fredd's number. It's not gregg's number 3) I've got basil's number. It's not gregg's number 4) I've got gregg's number. We have gregg's number". For a list of size [I]n[/I], we would expect, on average, for it to take half of the list size to find it, ie. [I]n/2[/I]. Quantum computers can use an algorithm known as [URL="https://en.wikipedia.org/wiki/Grover's_algorithm"]Grover's algorithm[/URL] that can find whether gregg is in this list at a speed of [I]sqrt{n}[/I]. The quantum computer is exponentially faster, this is obvious for large list sizes. I'm going to be tenuous with this card analogy so be warned, In the quantum situation, we print all of the list onto cards, do magic with them which we can't see ( there is no real world way to show this ), stack 'em, and then turn over the top card from the deck. What it actually would do, is instead treat all of the cards like a single special card. When it turns this card over, it knows there are specific probabilities of what the card may turn out to be - when the card is facing down, the card in theory could be any one of the cards as you haven't looked at it - the one with gregg's number on it, or george's number, etc. In the quantum world, the computer would first have to define a system to run in - this would be the decks of cards. Each card with a number would be assigned a "state". The turned over card can only ever be in one state. With the cards facing down, the top card is in a superposition - it could be any of the states as we haven't looked to see which state it is, and we've no way of knowing how it's been shuffled. We would quantum mechanically turn the system's state into a superposition by using a hadamard gate, this would also normalise the system such that each state would have an equal probability. The computer then does a sneaky check using a gate known as an "oracle"; this essentially is a binary "is the number gregg's number" check. Mathematically, imagine each state ( each possible card ) pointing in unique directions, these are now known as state [I]vectors[/I]. The system, that is the top card of the deck / card we will turn over, is also pointing in a direction. By how much the system vector points in the directions of these state vectors determines it's probability of, when observed (turned over), will turn out to be that particular state ( that number on the card ). The oracle gate, alongside a bunch of other gates ( diffusion stuff, a particular grover iteration), will cause the system vector to point slightly more towards the oracle state (gregg's number), and slightly away from the rest of the states. This means that the state is now more likely to, when observed, be the oracle state (gregg's number) than any other state. Surprisingly, the amount that doing this turns the system vector is actually a mathematically defined number. This means, that if you use the gates in a chain a certain amount of times, you will rotate the system vector so much so that it points exactly in the direction of the oracle vector - if you use the gates even more, it will start to point away! It turns out, that for a list of size N, you would have to repeat this iteration R times to point in the most optimal direction, where [t]https://wikimedia.org/api/rest_v1/media/math/render/svg/313d5618e49ca5f1b40236582973fe1519e37587[/t]. After doing this [I]r[/I] times, your system vector now points almost exactly in the direction of the oracle vector (gregg's numbers' card). If you were to now observe the system, ie. turn over the top card, you are statistically much more likely find it to be the card with gregg's number than any other, depending on how many times you iterated the search. Mathematically this is completely correct, however I'd have no idea how you would implement these gates to actually affect atoms or photons - but it's being done by IBM and such. Only certain things can be put into superpositions- I can't remember the exact criteria, but I believe it's if a quantity exists in a complex plane, such as the polarisation of light. [editline]30th October 2017[/editline] [QUOTE=Gmod4ever;52835135]Additionally, while the set of problems that [B]are[/B] massively parallelizable, to the point of being practical candidates for quantum solving, is respectably large; it is, at the same time, a [B]far[/B] smaller set of problems than people often like to think.[/QUOTE] Exactly - there are really only a few situations where these would even be necessary.

[QUOTE=thelurker1234;52834867]1s and 0s work just fine for that 0 because of that I think that the closest we'll get to having them in normal consumer computers is on a small chip dedicated to certain tasks that just make sense to do on a quantum computer. kind of like how AES encryption is done[/QUOTE] I think the video put it better; never on consumer devices but accessible online. If you're not going to use them for huge tasks, why is it necessary to have a dedicated chip on every device? [editline]29th October 2017[/editline] And the huge tasks they are good at are so goddamn large there is no way you're going to do them as a consumer.

I don't like the idea most analogies fall into about 'checking every possibility at once' and things like that, it makes people think it's some kind of magic multiverse computer thing which is always right in reality, the quantum logic is more like a sieve, which gradually filters out the incorrect answers. you still have to run it multiple times to achieve a high statistical separation for the winner but as instant mix said it's faster than the classical method - O(√n) instead of O(n) time

[QUOTE=krail9;52835898]I don't like the idea most analogies fall into about 'checking every possibility at once' and things like that, it makes people think it's some kind of magic multiverse computer thing which is always right in reality, the quantum logic is more like a sieve, which gradually filters out the incorrect answers. you still have to run it multiple times to achieve a high statistical separation for the winner but as instant mix said it's faster than the classical method - O(√n) instead of O(n) time[/QUOTE] You are correct about needing to measure the qubits multiple times in typical quantum algorithms, but I was talking specifically about this 4 card sorting example. The demo even specifically stated that only one operation was used to find the card. Given the simplicity of the problem and small dataset, it's highly likely that IBM didn't use a full implementation of Grover's algorithms, and one measurement is enough to determine the outcome.