Current State of Quantum Computing – Computerphile


Well, brief mentions in other videos about the Q-word, *Laughs* you know, the magic computers that are going to break all encryption and beat everyone at their own game and be the next big thing and be on all our smartphones. Should we put some reality on this? Where are we at with quantum computing? I Work at a company called Righetti computing. We’re a company in Berkeley, California And we build quantum computers from the ground up. We do the silicon fabrication all the way up to how we program the machines I in particular Think a lot about how we program the machines how we interface with it at some point Somebody has to type something into a keyboard to make it do something and that’s mostly where I think about things quantum computers Really should be thought of as a way to Augment our existing compute infrastructure exists in computers solve problems, and that’s not going away there. They will continue to solve good problems So a quantum computer really should be thought of as a is a special-purpose machine that can solve certain problems a lot better Than existing computers can solve them. In fact, I think thinking of it like a GPU is is pretty fruitful that GPUs are, are, they are computers in a sense. But really they’re they’re better thought of as like a coprocessor to your main computer. They GPUs are extremely good at solving a variety of problems you could sort of Wrangle a GPU to solve any problem you want to but it’s you know regular computers are good at that so quantum computers I think of as like on the side Unfortunately current quantum computers are gigantic. They’re they you know, they’re like the old computers like the ENIAC they fill a room But nonetheless you you hook your computer up to this and it does separate problem-solving as a company We build these quantum computers and deploy them there. They’re Accessible right now on the cloud. There’s something that I on my laptop anywhere Actually, I don’t need any special interface can connect up to them and do computations with them And we’re at a point now where? the fundamental unit of resource of a quantum computer just as we take regular computers and think about them in terms of bits and gigabytes and so on on a quantum computer These are measured in qubits, which are very special because every time you add a qubit You’re you’re doubling the capacity of the computer in some way. So Last year, we released a machine that has around eight qubits. And now the current machine we have is about nineteen qubits so it’s not that we Approximately just doubled the qubit because each qubit itself is doubling so you have to think of that number of doublings each time from eight to nineteen problems that deal with lots of interactions or lots of Possibilities so to speak are where quantum computer sort of shine so one big area of application is in quantum computing is the simulation of molecules molecules are are made up of a bunch of atoms and each of these atoms is is Applying a force to the other atoms they’re pushing and pulling each other and they’re just lots of ways in which the atoms interact with one another and a quantum computer is extremely good at Keeping track of all that and dealing with that in a very efficient way that classical computers which are normal computers Don’t deal with quite as well Likewise other types of problems like optimization which in mathematics means kind of finding the best or worst possible solution to a problem Quantum computers are showing a lot of promise and with that said though There’s one point I want to make which is sort of in popular science people think that quantum computers just sort of try every possibility At the same time you get all these sort of kooky Interpretations of multiple universes and so on and quantum computers, don’t do that. They do not try every possibility at the same time But the second point is that applications of quantum computing is a very active field of research quantum computers currently are They’re called noisy quantum computers. They they sort of act like analog devices not digital devices And so there’s all this a little bits of error and little bits of noise that come into the system and so it’s an active field of research to see what Algorithms and what problems are very robust to this noise and it turns out that this molecular simulation is an example of a problem That’s very robust to noise Whereas what? You might heard of like Shor’s algorithm and factoring and breaking encryption are not robust annoys at all They’re very difficult to to do so that’s a very active field of research both at the company that I work at As well as more broadly the community the sort of room size computers are sort of where we’re at the way in which these computers have been constructed is Using sort of off-the-shelf components that have not been specially made for the construction of that machine in general So you find generally with these quantum computers you find big racks of analog electronics these electronics that that have existed for other applications and so It’s definitely not at a point where we’ve we’ve sort of custom fabricated the entirety of the machine I’m not just talking about the quantum chip itself But the thing that the quantum chip is housed in the electronics that go with the quantum chip, etc, etc We’re only just starting We just in general as a in the field of quantum computation are only just starting to make like customized electronics and the like for these systems The other aspect of this is that just like the ENIAC for example Which had vacuum tubes that burnt out and you had to go replace them Literal bugs getting in there and so on neither of those things happen with quantum computers But what does happen is that the system in general isn’t shown to be stable across, you know? Five or ten years and just hasn’t existed that long And so components are are often swapped out often changed It’s not robust to changes in the environment like if somebody walks by with a big magnet the whole thing kind of goes awry, so it’s a really sort of Finicky machine not unlike the early computers like the ENIAC and so on How far are we off quantum smartphones? Yeah, that is that is really really far away because we need to cool these chips down to like the same temperature or colder than The temperature of outer space so you need big refrigerators as they’re called Not going to happen for a good while The interesting thing though is When we have multiple qubits, and this is really where the power of quantum computation happens We can actually think of it sort of simply Diagrammatically that if qubits if I just represent them sort of as a circle here, maybe we have three the idea Is that these qubits can interact this guy can interact with this guy?

54 thoughts on “Current State of Quantum Computing – Computerphile

  1. I still agree with Einstein about quantum physics. I think these computers are operating by something other than "quantum" physics. I think they are simply attributing something to quantum that is not.

  2. How is biology based (Muscle cell grid — actin and myosin) parallel computing (biocomputers) advancing? And could biological computers offer any kind of competition to the quantum computing that you’re working on now or are they too dissimilar?

  3. I don't trust QC. What if you need to turn it off and on again, but it is already on and off at the same time?

  4. Hasn't D-wave released a machine, D-Wave 2000Q, with 2000 qubits (with only 1000 qubits enabled)?
    What am I not understanding with this machine and their current machine being 19 qubits – surely they would know about a 2000Qubit computer so what am I missing?

  5. A quantum computer is a fascinating thing… – Will it blend? Or, which wins, a quantum computer or the Hydraulic Press?

  6. So a quantum computer is a special purpose computer eh? kinda like the first computers were I suppose. I wouldn't be surprised if we had quantum computers in our pockets in 20 years or so

  7. I wonder if time crystals might show promise for making quantum computing happen at a higher temperature?

  8. I'm pretty sure smartphones already use cloud computing services regularly, and quantum cloud computing will probably be all that there is in the foreseeable future.

  9. 19 Qubits, if "it" doubles with every qubit 2^19 is about 500.000. But what is "it"? Is it storage? Processing power?

  10. Is it future connecting your blockchain based app to quantum servers to do the math and return you output? This requires high network speeds but we do have them…

  11. Plot Twist
    This video is there just to fool the general public and to make them believe that Quantum Computers are still at their infancy.

    They are probably breaking all sorts of encryption with quantum computers, they just don't want to reveal it.

    Remember the imitation game ?

  12. I wonder if David Baker's group is working on this? Protein crystallography is rapidly becoming obsolete due to advances in cryo electron microscopy. Cryo EM may soon be a redundant as well, once we can brute force protein folding with a quantum computer.

  13. Yeah, well I "work" for Quantum Silicon in Alberta, and we'll have 500 qbits in a year. Room temp. Birches

  14. I notice that I don’t use the letter Q in my life as much as I now realize that I have been needing to.

    Positively life changing

  15. I would have liked some discussion on D-Wave's quantum computer. Aren't they planning on making a 2000 qubit quantum computer? Seems to me that D-Wave is worthy to be discussed.

  16. What does the data set you feed a quantum device with look like? I mean, is there a math treatment like a DAC between the two interfaces or can it deal with binary logic just as well as the electronic ones?

  17. can someone expand on the robustness to noise issue he mentioned? I remember hearing in another video, maybe a computerphile video, that there are two types of quantum computer… one of them gives correct and consistent answers every time, while the other one is somehow not a "true" quantum computer because the process is somehow subject to error and inconsistency. So you don't always get correct answers. Would that be the noise they're referring to? and would that mean that the type of quantum computer he's referring to cannot be used for problems that need a single specific and accurate answer like cryptographic problems? are we ever going to reach that level of computing?

  18. The current state of quantum computing is a superposition. There's no way to know with certainty where it will be going.

  19. 😇 Thank you for all your beneficial video, it certainly is greatly valued and I definitely value your hard work !👍

  20. well the title of this video is Current State of Quantum Computing not Introductory Course to Quantum Computing isn't it? He actually answered the question. I like this guy. He sounds like my (American) professors and makes a lot of references to industry.

  21. Why are they still so big they should be able to start making them smaller thanks to higher-order topological insulator and time crystals. These things fix the high temp super conductor problem and coherence problems respectively

  22. Its scary…. computers in the 70's were size of a room, they ran off binary. Look at where we are now with binary computers.. In 40 years time we will have quantium computers inside of what we call a laptop now.

  23. So if quantum CPUs don't conceptually "multi-core" (to a power rather than linearly) with additional qubits, then how exactly do QPUs work? I know roughly how they work, quantum logic gates, superpositions of superpositions, entanglement, interference and so on. I'm also familiar with applications such as optimization (traveling salesman) and modeling of quantum, or natural systems (molecules, fundamental particles, etc). My background is physics/finance not computer science. From a physics POV I think of superposition and especially superposition of superpositions (e.g. young's double slit) as implying multiple processes simultaneously – so the electron or photon exhibits wave-particle duality and in some sense travels all possible paths before deciding upon a definite outcome upon observation, an outcome that is influenced by past and future events as well as the present (evidenced by the clearly defined interference patterns). This is analogous to qubits being 0 and 1 at the same time (a superposition) but that means the value could be anything between 0 and 1 until we measure it. Does adding qubits not allow the system to solve branching math problems by branching out probabilities (superpositions of superpositions) in a similar way to an electron in young's double slit? That is to say when you ask a QPU to perform an action the qubits in a sense know the answer and all possible answers to the question upon input, the perceived delay is just how long it takes the observer (measurement, output) to experience the output in their reality.

  24. What cryptographic protocol is best suited for IoT? Directed Acyclic Graphs? Hash tables? Blockchains?

  25. If sciense fabricate a new material that could allow a new CPU with the same advadages BUT to work in normal computers temperature we will have Quantum computers.Although a very importand question is not asked so far.If we can create a so powerfull CPU, what could happen if we overclock it ?

  26. Isn't a quantum computer the same as a regular computer but instead using the spins of say a photon to represent a bit unlike conventional computers which use a voltage?

  27. see what he says at 3min , well people that is recording a very fast RANDOM process THAT IS NOT COMPUTING .
    THE QUANTUM BY DEFINITION ONLY DOES RANDOM ! DONT YOU CRETINS GET IT , THE QUANTUM BIT CANT DO THE ACTUAL COMPUTING ….. YOU TOTAL CRETINS .

Leave a Reply

Your email address will not be published. Required fields are marked *