First, I would like to thank D-Wave and Dr Dominic Walliman for putting together these videos. I work in high-performance computing, and watching them attentively I think I have a good understanding of how adiabatic quantum computers can tackle some currently intractable problems efficiently. I have a question: my understanding is that the actual optimization problem we are trying to solve is encoded in the biases and the couplers. Do you happen to have resources to get a better understanding of how to translate the initial problem into this configuration? How can a programmer get a sense of that ? What languages and APIs are available at the moment?
It is very correct description. Understanding quantum wave theory is required to see the valleys are the bottom of the wave which is how particles are found in a position. To see the quantum effect we must first understand the bias spoken of is merely the machine applying a slight magnetic field which hinders the chaos phenomenon and installs a desired effect which with minimum influence allows a known position to be found. Annealing is what qualifies this system as a quantum system which if quantum entanglement is true then so is this. However quantum entanglement is like gravity. We know there is an effect but we really cant prove it exist but we know it does to a degree. As long as the computer works then is a true quantum computer relatively speaking.
yeh but nothing has really been explained. materials to do this etc what is the "best" answer. traditional computers its either on or off and then you can make apps switch into different states or bad chess moves not be valid unless its the only option. but this doesnt explain anything
The rub is assigning values to the biases and couplers, it is the same as a horse racing handy-capping spreadsheet, the rub is assigning weighted values and offsets.
Can you correct me if i am wrong: Quantum tunneling is the switch from one quantum state to another which would classically require external energy to overcome the "hill" between the states, but "tunneling" describes the phenomenon that this external field/energy is not required and the state just "jumps" from one to another?
I think I'm finally understanding how a quantum computer actually works, but I would like an applied example like the Shor's algorithm or the Grover's algorithm.
+Mackaber Witckin They explained in a previous video that this form of quantum computing is not (in my phrasing) 'Quantum Computing Complete'. It can't actually run Shor's algorithm.
+Mackaber Witckin An actually example would be "The Traveling salesman problem" or process scheduling problem. Other chaos problems that classical computers have a really hard time with Weather prediction, machine visualization, machine learning would also be in the subset of problems that quantum computers are good at.
rRobert Smith as i understand it, basically all optimization problems with a finite solution space.. But if this really outperforms classical optimization algorithms like hill-runner / watershed by a large margin, it will solve a lot of problems!
5:32 which large quantum object you are talking about? and who changes the probability ? The large quantum objects or the programmer? How the large quantum object knows what biases and coupling is to be applied?
Could this be a modifier for dealing with A.I. Personality? Is it stable to use? Also since the values exponential increase stabilized to doubling to the nth would it be too predictable in cases of encryption?
This sounds exactly the same as adiabatic quantum computing, but according to wikipedia, adiabatic quantum computing is a subclass of quantum annealing - so how does quantum annealing differ from adiabatic quantum computing?
Now i understand :) thanks for the really great video! I think it really depends on the number of qubits we are able to entangle in the end and how long a larger number of qubits take to reach their "stable" low energy state in the end..
What level of bias can be put on a single qubit. Could you affect it enough to have reasonable certainty of a 1 or a 0 from one qubit or would you have to use several in conjunction?
it sounds to me like a sorting machine, a sort of custom made filter for each task, with the purpose of identifying a specific set of variations, in the end its still all about conventions
Why use a quantum computer with its size, and very specialized required conditions(in the millikelvin) when there is an optical computer much smaller that is capable of more than an exaflop of compute?
Remember, there's no "forcing." It's just defining energy levels and then trying to minimize among them. The highest energy states would be 000 and 111, whereas 001 (two pairs different and one the same) would have the same (lower) energy as 010, 011, 100, 101, and 110. So it would equally prefer any of those 6.
It uses the principle of quantum physics, where a quantum state is in all possible states at the same time. Only once we measure it, it collapses to a single state. Google Schroedingers cat and double slit experiment :)
It seems like the x-axis is the magnetic field direction. The end states have the same y-axis value but different x values. The end states are also either 'North Up' or 'North Down'.
+Thor Mentha The main driving force behind research into quantum computing is Cryptography. Because of the fragile nature of entanglement & how measurement, irreversibly, and decoherence in quantum mechanics all work; attempting to intercept/decipher a message that has been encoded using a quantum computer would destroy the message altogether (and tip off the receiving party, since it is physically impossible to copy the states of a quantum system). So basically, having 2 quantum computers would grant a secure line of encrypted communication that is impossible to break. On the other side of that, quantum computers would make short work of any type of encryption used by a standard computer today. This makes it a very appealing research area for government agencies.
+Thor Mentha It would seem like a useful tool in artificial intelligence as well. Especially in deep neural networks and machine learning systems that require heavy processing power for statistical computation (it seems to me that quantum computing by design relies A TON on probability statistics).
+Thor Mentha In molecular modeling, we are almost always running optimization(annealing) problem, where drug molecule/proteins docking in to important proteins in our body. The optimizations are almost too slow to be useful on even state of the art classic computer (goes exponentially with size of protein). Q-annealing can solve that problem almost linearly, allowing much larger protein systems to be calculated. It will be game changing in drug development (so called in silico screen), chemical biology and even change the whole theoretical chemistry.
+Thor Mentha Optimization problems are all over the engineering field. For example, maybe you are trying to find the shape of an airplane wing which causes the least amount of drag with the greatest amount of risk. You can build a simulation to try numerous wing shapes but it would take too long to try enough of them using a conventional computer.
Isn't that exactly what people do to overclock classic computers? quantum computers consume no energy and have no heat dissipation though. I mean, the reasons for cooling are totally different.
No, you cool the computer because otherwise the processor will just start to melt because of the increase energy consumption which needs to get dissipated
so with the new computer with 1000 qubits that's well over ten novemnonagintillion (1 x10^301) different possible values... that's a lot more then modern "classic computer" processors I know of since it think most 64x based ones are around eighteen quintillion? Also has any of this put to any practical use yet or is it still all at point of experimentation or theoretical work? Also I apologize if any of my phrasing is off I'm just a high school student completing a report
we cant control them currently we can only control the final and the initial state through quantum annealing we still have to achieve a gate model quantum computer
For a given couple, there are surely only two states though? The coupler has already decided the coupled qubits' behaviour to be same or different, and thereafter can only be in two states, surely. If couple opposite they can only be 10 or 01 ; if same, 11 or 00.
I wonder if D-Wave Systems will ever reply to this question, but would that be the end of internet security? I wonder how easy would it be for a D-Wave computer to bruteforce and deactivate, say... homeland security nuclear defence umbrellas to give an example...
u can brute force of they wont allow u to enter the password more than a few times
7 років тому
By internet security do you mean Authentication (logging in), or Encryption? From what I understand D-Wave's quauntum computers can not break encryption. They are not designed to solve those types of problems, only optimization problems. As for authentication, well it depends if you are using a Key base scheme to log in and authenticate. If its a simple password to log in, then quantum computers can't help you there.
Ok, so thats the theory, bit confused in why the lowest energy state is the best answer. and what is an answer and how is the best answer decided anyway. also how do you even build something at this micros level. i can actually get the idea of on and off and shrinking it into a chip using the right materials but nothing here has been explained in how to put this into even a large chip. its very suspicious
«Лучше быть умным и иногда тупить, чем быть тупым и умничать». Зачастую мы хотим произвести хорошее впечатление. Это можно сделать не только красиво и стильно одевшись, но и своей грамотной речью и своим лексиконом. Ведь не даром говорят: «Встречают по одежке, а провожают по уму». Чтобы произвести то впечатление, которое вам нужно, достаточно выучить хотя бы 50 умных слов с их значениями.
entirely agree, this is pseudo superposition, mimicking quantum fundamentals to perform calculations, true entanglement would indeed make you god on earth #scaryHuh
What a great explanation of a nonsensical concept. It's a quantum explanation... an explanation and non explanation at the same time. Seems appropriate. An exercise in techno babel. No one want to appear less intelligent, so they pretend to understand and try to parrot back what they were told... Quantum entangled 1-1, 0-0, 1-0, 0-1...with each additional Q-bit the calculations double... both states at the same time. Got it... Now I can sound smart to the other pretend geniuses who are trying to impress me with their knowledge of Quantum Mechanics. The real truth behind the D-Wave computer is this... It works for the same reason that an Ouija board works. Study it and understand both !!!
I don't believe true IBM large scale quantum computers are possible. This, however, seems like it may be, and, while limited, is good at solving optimization problems, which will be greatly needed in the near future.
Quantum computers and computing is a scam and a very expensive scam. Just the simple thought that a quantumbit can be both "one af zeo" at the same time is false, it can only assume one state at one single timestate - however small the timestate is - not two state in the same timeslice but only one state. Now lets assume it's possible to have both state at the same time, - them two qbit would represent 4 different state at the same time - but only state is right(correct). That means there is 75 % chance it's a wrong result. Adding quantumbits makes it just even worse - 3 qbits gives you 6 different states - but only one is correct equals 16.66% and the other 83.33 % will be wrong. OK lets calculate with "the speed of light" with quantumbits as being predicted - the quantum computer will be able to solve problems that wil take 10.000 years for ordinary computers to solve but wil be solved i a second with a quantumcomputer. - How wil you control and validate the result to be sure its the right result - and not wrong? - By the way th ecompagny D-wave is selling "quantum computers" and have done i forsom time already, - but where is all the fantastic result these computers are producing - it can't all be secret calculation when one bear in mind that some of these computers have been sold til commercial compagnies.
First, I would like to thank D-Wave and Dr Dominic Walliman for putting together these videos.
I work in high-performance computing, and watching them attentively I think I have a good understanding of how adiabatic quantum computers can tackle some currently intractable problems efficiently.
I have a question: my understanding is that the actual optimization problem we are trying to solve is encoded in the biases and the couplers. Do you happen to have resources to get a better understanding of how to translate the initial problem into this configuration? How can a programmer get a sense of that ? What languages and APIs are available at the moment?
A very clear explanation of a very complex subject. The examples were very relatable.
It is very correct description. Understanding quantum wave theory is required to see the valleys are the bottom of the wave which is how particles are found in a position. To see the quantum effect we must first understand the bias spoken of is merely the machine applying a slight magnetic field which hinders the chaos phenomenon and installs a desired effect which with minimum influence allows a known position to be found. Annealing is what qualifies this system as a quantum system which if quantum entanglement is true then so is this. However quantum entanglement is like gravity. We know there is an effect but we really cant prove it exist but we know it does to a degree. As long as the computer works then is a true quantum computer relatively speaking.
Beautifully explained
Companies like you, Makes lessons in our syllabus for learning............Who are happy or sad (Me included)
This is revolutionary! The possibilities are endless with quantum computing!
It's nice they admit this doesn't do everything a full blown quantum computer can do. Clears up a lot of confusion.
"and all of this happens in about..twenty microseconds.."
I really like the simple explanation for a complexe mechanism.
yeh but nothing has really been explained. materials to do this etc what is the "best" answer. traditional computers its either on or off
and then you can make apps switch into different states or bad chess moves not be valid unless its the only option. but this doesnt explain anything
The rub is assigning values to the biases and couplers, it is the same as a horse racing handy-capping spreadsheet, the rub is assigning weighted values and offsets.
whats a coupler made of ?
Can you correct me if i am wrong: Quantum tunneling is the switch from one quantum state to another which would classically require external energy to overcome the "hill" between the states, but "tunneling" describes the phenomenon that this external field/energy is not required and the state just "jumps" from one to another?
I think I'm finally understanding how a quantum computer actually works, but I would like an applied example like the Shor's algorithm or the Grover's algorithm.
+Mackaber Witckin They explained in a previous video that this form of quantum computing is not (in my phrasing) 'Quantum Computing Complete'. It can't actually run Shor's algorithm.
+Mackaber Witckin An actually example would be "The Traveling salesman problem" or process scheduling problem. Other chaos problems that classical computers have a really hard time with Weather prediction, machine visualization, machine learning would also be in the subset of problems that quantum computers are good at.
rRobert Smith as i understand it, basically all optimization problems with a finite solution space.. But if this really outperforms classical optimization algorithms like hill-runner / watershed by a large margin, it will solve a lot of problems!
Great and the animations are spot on. Could you also take a simple heuristic optimisation problem and explain how the annealing process works
5:32 which large quantum object you are talking about?
and who changes the probability ?
The large quantum objects or the programmer?
How the large quantum object knows what biases and coupling is to be applied?
its really weird and suspicious, this physics was all theory that its based on as well
Is that Dominic Walliman from the channel DoS (Domain of Science)?
Edit: Checked description. It's him.
4:25 Would it work to have 3 straight cuplers and 1 twisted, or is that impossible?
I imagine it becomes like a Mobius loop.
(01)-0-0-(01)~ or (01)-1-1-(01)~ ...n•••
Could this be a modifier for dealing with A.I. Personality? Is it stable to use? Also since the values exponential increase stabilized to doubling to the nth would it be too predictable in cases of encryption?
This sounds exactly the same as adiabatic quantum computing, but according to wikipedia, adiabatic quantum computing is a subclass of quantum annealing - so how does quantum annealing differ from adiabatic quantum computing?
1:10 How do we create this barrier ? With a Hadamard gate ?
Now i understand :) thanks for the really great video! I think it really depends on the number of qubits we are able to entangle in the end and how long a larger number of qubits take to reach their "stable" low energy state in the end..
What level of bias can be put on a single qubit. Could you affect it enough to have reasonable certainty of a 1 or a 0 from one qubit or would you have to use several in conjunction?
but can it run Mad Max on ultra settings ?
Was it Feynman who proposed the idea initially?
That was a great explanation. awesome... Thanks,
What is a coupler physically ?
The external bias seems to be important. How is the external bias controlled?
it sounds to me like a sorting machine, a sort of custom made filter for each task, with the purpose of identifying a specific set of variations, in the end its still all about conventions
Why use a quantum computer with its size, and very specialized required conditions(in the millikelvin) when there is an optical computer much smaller that is capable of more than an exaflop of compute?
***** yes but brute force can accomplish the same task, and faster too.
Where'd you go to school just wonderin g if you're the joe mills i know
+Alex xeon when u have the problem being solved thru multiple streams of quantum computing u can find the solution instantaneously
Oooo! SLAMDOWN B)
What happen if you force an impossible situation? like reverse entangle A with B, reverse entangle B with C and reverse entangle C with A?
Remember, there's no "forcing." It's just defining energy levels and then trying to minimize among them. The highest energy states would be 000 and 111, whereas 001 (two pairs different and one the same) would have the same (lower) energy as 010, 011, 100, 101, and 110. So it would equally prefer any of those 6.
soon they will have dwave mini
how it stores the both states of ( 0 or 1 ) at the same time .
By superposition I think :)
It uses the principle of quantum physics, where a quantum state is in all possible states at the same time. Only once we measure it, it collapses to a single state. Google Schroedingers cat and double slit experiment :)
What does x-axis represent in the graph 1:27?
It seems like the x-axis is the magnetic field direction. The end states have the same y-axis value but different x values. The end states are also either 'North Up' or 'North Down'.
You need to make more videos
Probably the least significant detail of this video, but I really like D-Wave's jingle at the beginning.
That smooth bass
I like it too!
What are the applications of this? Making topographic maps?
moniker127 Nobel Prize then?
+Thor Mentha The main driving force behind research into quantum computing is Cryptography. Because of the fragile nature of entanglement & how measurement, irreversibly, and decoherence in quantum mechanics all work; attempting to intercept/decipher a message that has been encoded using a quantum computer would destroy the message altogether (and tip off the receiving party, since it is physically impossible to copy the states of a quantum system). So basically, having 2 quantum computers would grant a secure line of encrypted communication that is impossible to break. On the other side of that, quantum computers would make short work of any type of encryption used by a standard computer today. This makes it a very appealing research area for government agencies.
+Thor Mentha It would seem like a useful tool in artificial intelligence as well. Especially in deep neural networks and machine learning systems that require heavy processing power for statistical computation (it seems to me that quantum computing by design relies A TON on probability statistics).
+Thor Mentha In molecular modeling, we are almost always running optimization(annealing) problem, where drug molecule/proteins docking in to important proteins in our body. The optimizations are almost too slow to be useful on even state of the art classic computer (goes exponentially with size of protein). Q-annealing can solve that problem almost linearly, allowing much larger protein systems to be calculated. It will be game changing in drug development (so called in silico screen), chemical biology and even change the whole theoretical chemistry.
+Thor Mentha Optimization problems are all over the engineering field. For example, maybe you are trying to find the shape of an airplane wing which causes the least amount of drag with the greatest amount of risk. You can build a simulation to try numerous wing shapes but it would take too long to try enough of them using a conventional computer.
Test it in rendering Please !
if we supercool a classical computer will its performance increase?
+Amit saxena Yes, less resistance.
Isn't that exactly what people do to overclock classic computers? quantum computers consume no energy and have no heat dissipation though. I mean, the reasons for cooling are totally different.
No, you cool the computer because otherwise the processor will just start to melt because of the increase energy consumption which needs to get dissipated
so with the new computer with 1000 qubits that's well over ten novemnonagintillion (1 x10^301) different possible values... that's a lot more then modern "classic computer" processors I know of since it think most 64x based ones are around eighteen quintillion? Also has any of this put to any practical use yet or is it still all at point of experimentation or theoretical work?
Also I apologize if any of my phrasing is off I'm just a high school student completing a report
we cant control them currently we can only control the final and the initial state through quantum annealing we still have to achieve a gate model quantum computer
For a given couple, there are surely only two states though? The coupler has already decided the coupled qubits' behaviour to be same or different, and thereafter can only be in two states, surely. If couple opposite they can only be 10 or 01 ; if same, 11 or 00.
Oh nevermind... I should have kept listening. The coupling is also probabilistic, so it makes sense now
I just paused the video to say, love you hair dude.
thanks
Good for AR Calculation
Alright mate!
I wonder if D-Wave Systems will ever reply to this question, but would that be the end of internet security?
I wonder how easy would it be for a D-Wave computer to bruteforce and deactivate, say... homeland security nuclear defence umbrellas to give an example...
u can brute force of they wont allow u to enter the password more than a few times
By internet security do you mean Authentication (logging in), or Encryption? From what I understand D-Wave's quauntum computers can not break encryption. They are not designed to solve those types of problems, only optimization problems.
As for authentication, well it depends if you are using a Key base scheme to log in and authenticate. If its a simple password to log in, then quantum computers can't help you there.
Ok, so thats the theory, bit confused in why the lowest energy state is the best answer.
and what is an answer and how is the best answer decided anyway.
also how do you even build something at this micros level.
i can actually get the idea of on and off and shrinking it into a chip using the right materials but nothing here has been explained in how to put this into even a large chip.
its very suspicious
«Лучше быть умным и иногда тупить, чем быть тупым и умничать».
Зачастую мы хотим произвести хорошее впечатление. Это можно сделать не только красиво и стильно одевшись, но и своей грамотной речью и своим лексиконом. Ведь не даром говорят: «Встречают по одежке, а провожают по уму». Чтобы произвести то впечатление, которое вам нужно, достаточно выучить хотя бы 50 умных слов с их значениями.
This is fuzzy logic derived via quantum mechanical imitation, not a true quantum computer
Cool so qubits run like an A/C current and it can be influenced like a slot machine being subjected to a magnet?
Not real actual entanglement (that would be machine of God century), this new design appears to be simply analog statistical computing
entirely agree, this is pseudo superposition, mimicking quantum fundamentals to perform calculations, true entanglement would indeed make you god on earth #scaryHuh
yes magnetism and entanglement are the same thing. see www.quora.com/Is-magnetism-a-quantum-phenomenon
What a great explanation of a nonsensical concept.
It's a quantum explanation... an explanation and non explanation at the same time. Seems appropriate.
An exercise in techno babel.
No one want to appear less intelligent, so they pretend to understand and try to parrot back what they were told...
Quantum entangled 1-1, 0-0, 1-0, 0-1...with each additional Q-bit the calculations double... both states at the same time.
Got it... Now I can sound smart to the other pretend geniuses who are trying to impress me with their knowledge of Quantum Mechanics.
The real truth behind the D-Wave computer is this... It works for the same reason that an Ouija board works.
Study it and understand both !!!
Could you pass me some info or links?
So you’re in my area. It’s bs. It does something tho
I bet Schrödinger would love seeing this
This is way over my head.
into zer darkness, we are one!!!
what are they doing with these computers? running minecraft benchmarks?
And the second is bruteforcing every single defence system. Like a nuclear defence umbrella.
20 microseconds is actually quite a bit of time, surprising
I don't believe true IBM large scale quantum computers are possible. This, however, seems like it may be, and, while limited, is good at solving optimization problems, which will be greatly needed in the near future.
Ok, but what you have described is actually a classical computer.
Ok now I am scientist...
skynet system,,,
PC QUANTUM MASTER RACE !!!!
:P
Playing GTA 6 with two dwave cards in SLI would be awesome.
I was thinking about doing that with Minesweeper, or solitaire.
Ok, so it went from fantasy, to defining bits to bytes.
Ether torsion.
no such thing as besx or infx or not, infx, do any nmw
This is ge=reat!
Cubits...Qubits... Cube it's....
Ok
Quantum computers and computing is a scam and a very expensive scam. Just the simple thought that a quantumbit can be both "one af zeo" at the same time is false, it can only assume one state at one single timestate - however small the timestate is - not two state in the same timeslice but only one state. Now lets assume it's possible to have both state at the same time, - them two qbit would represent 4 different state at the same time - but only state is right(correct). That means there is 75 % chance it's a wrong result. Adding quantumbits makes it just even worse - 3 qbits gives you 6 different states - but only one is correct equals 16.66% and the other 83.33 % will be wrong. OK lets calculate with "the speed of light" with quantumbits as being predicted - the quantum computer will be able to solve problems that wil take 10.000 years for ordinary computers to solve but wil be solved i a second with a quantumcomputer. - How wil you control and validate the result to be sure its the right result - and not wrong? - By the way th ecompagny D-wave is selling "quantum computers" and have done i forsom time already, - but where is all the fantastic result these computers are producing - it can't all be secret calculation when one bear in mind that some of these computers have been sold til commercial compagnies.
Mandela effect
whats has this to do with that ?
Sumit, you will probably never find out... Thats the sad part :( They dont tell you...