Quantum computer for 50 qubits.

Currently, every self-respecting firm is trying to create a quantum computer. Such as Microsoft, Google, NASA, Russian Quantum Center and others. Some, until no one understood the essence of the matter, are already making money on these quantum computers. For example, D-Wave Systems managed to sell their first 2000-qubit quantum computer for 15 million dollars. At the same time, Googol is trying to build a 50-qubit computer, but so far it has failed. Others are not doing better either. The program manager at Google is John Martinis. He gave us a lecture on the achievements in this area. She is recorded in the video "ICQT 2017. John Martinis, Google: The Quantum Computer: Life After Moore's Law" .

I tried to understand what John is saying and to express his thoughts in my own free interpretation. I am aware that many will understand it deeper and more correctly and, perhaps, will also tell about it.

John began with a story about how he explains the basics of quantum mechanics to his children. First, he explained to them, "that the electron has such an indefinite structure, that is, it is, as it were, fuzzy, blurry" . In English, such a structure is denoted by the word phasie. This means that "the electron has such a nature that it can be randomly distributed around the nucleus. And what is very interesting is this randomness of quantum physics, this is a picture that is difficult to describe. However, how the electron is distributed in its orbit around the hydrogen nucleus is a very exact science. You can very accurately describe what exactly happens to this electron over time. Thanks to this, we can build computers. Without this, we could not build computers. This is a very exact science. "

Naturally, it makes no sense to have a discussion with a generally recognized scientist, although it's funny. It is difficult to understand how an electron distributed randomly around a nucleus can be described very accurately. What can we accurately describe in his behavior over time? What are we? can we accurately determine the coordinates of its movement? Can not. Heisenberg, Schrodinger and the heat movement forbade us to do this. True, the latter is fixable. Schrodinger, on the other hand, turned the corpuscle into a wave and blurred it, or, as dear John says, gave it the status of a phasie. Nobody knows what point of the wave and how to measure the spatial position of this point. Yes, in fact, they do not know anything about the wave. It has no amplitude, no overall length, no thickness, no mass, nothing physical at all. It's just an abstract symbol.

What else can be described in the behavior of an electron in time? Changing its size, mass or charge is also impossible, because it is a wave that simply indicates the probability of finding a particle, like a wave in some place. And a particle is a wave, and its location is a wave. In this case, it is somehow automatically assumed that the distribution of the electron is precisely or very accurately described relative to the nucleus, and it is also forgotten that the nucleus itself is also described by the Schrodinger wave function and its location is likewise probabilistic. I'm not talking about protons, quarks, atoms.

It is clear that John can recognize all this reasoning as a classic that does not work in the quantum world. And he will be sure that he is right. And why? And how will his children respond to such reasoning? Exactly the same as he. There are a lot of such people now. You see on the net how many young talents boldly talk about quarks, Jung's experience, the theory of relativity and everything else. They picked up this from the older generation, and the older generation, at the suggestion of Schrodinger and some other "theoreticians", believed in the queen of sciences mathematics and gave up the search for strength in phenomena through philosophy and experience, and took up the pen. But the queen turned out to be the Gorgon Medusa, and science has turned to stone for a hundred years.

What's left? There is only one thing left - spin, that is, the magnetic component of the electron. They try to work with him, but they also try to blur him.

John further states that "an electron near the nucleus emits waves. And it is something like these stationary waves ... and in quantum mechanics, different states of the electron are possible ... here you see the electron's orbit, this is the state of rest of the electron. The state in which the electron naturally finds itself. However, the electron may still have an excited state. And we can use these two states, the resting state, the excited state, for calculations. And we can develop a system that simultaneously calculates both of these cases, when the electron is at rest and in an excited state. And instead of classical computers, which consider one example, then another example, you can develop a computer that takes into account these two states at the same time. This is called the state of the qubit ".

Some kind of porridge. Maybe this is the translation, or John doesn't know what he is talking about. As soon as Rutherford presented his planetary model of the atom, everyone began to argue that an emitting electron could not stay in a stationary orbit. He must definitely fall to the core. And very quickly. But then Bohr appeared and said that in a stationary orbit, no matter what: excited or basic, the electron does not radiate and therefore does not fall on the nucleus. And radiation or absorption occurs only during the transition from one state to another. And there are a lot of these states. Look at the spectrum of hydrogen. And they are all at least slightly, but differ in energy. As a satellite that can fly at an altitude of 100 kilometers or an altitude of 100 kilometers and 3 meters. But for John, only two states of the electron are enough. He simply calculates these two states at the same time. At this point in his story, John opened the eyes of the Russian public that there should be little emphasis on the idea that a qubit can have many different states and that they can be simultaneously processed. Just two states are enough. The focus shifts to the number of two-state qubits.

This is something like this. Let's take some lengthy computation, for example, extracting some root from a large number. Then the computation itself will take place in the processor itself, the calculator. And service programs will take little time. Let the concrete computation of the root be 100 seconds and the serving 2 seconds. Let's take the next problem that is approximately the same. Then an ordinary computer will actually complete them in about 204 seconds. If your computer completes these two tasks in 102 seconds, provided that it completes each task separately in the same time, then the flag is in your hands. Indeed, the calculations are running in parallel.

Then John explains the effect of concurrent computing in even more detail. He "is very interested that an electron can be at any point in its orbit at the same time, even at different points at the same time. And what is interesting is that an electron can be simultaneously at rest and in an excited state, which from the point of view of the classical laws of physics, it simply does not make sense, but this is a feature of quantum physics.

Thanks to this, we can perform calculations using such an atom. And you see at the bottom, I used parentheses to represent these two states: 0 is a resting state, and it says plus one - this is an excited state. This means that the electron is in two states at once. And why is it interesting. This means that with one qubit calculation, we may not separately calculate zero and receive answers, one - to receive an answer. We can simultaneously calculate some problem at the same time and at zero and one. In parallel, computing is taking place. And this is great because this quantum computation allows you to calculate two cases at the same time. This is twice as fast as if using a classic computer. This is great ".

Take a look: an electron can, I highlight the word can, “" is at any point of the orbit at the same time, even at different points at the same time ". So how is it really: maybe it is, or maybe not? Is it the choice of the electron itself? So maybe the electron itself chooses the number of points themselves and their location? After all, you need to be guided by something when you try to accurately or even very accurately describe the behavior of an electron in time.

Despite the peculiarities of quantum physics, as John understands it, I still want to ask, to clarify: is this the same electron in three places or are these three identical electrons? That is, it is three masses or one? If an electron has such a wonderful property of being in different places at the same time, why shouldn't it acquire a lot of masses? Nonsense from a classical standpoint, but from a quantum standpoint it can be.

Of course, it is impossible to get any clear answer to these questions, but John confidently adheres to the position of superposition in this fragment of his speech and further develops it. This is that we only need two states from a qubit. The main thing is that they are simultaneous, and then we will process them.

This is what will blow us away. ”But you might say: well, this is all very complicated, and we are only doubling our computing power. But now we have two qubits. Each of them is in this dual state: zero plus one. And together, these qubits can have 4 states that occur simultaneously. And now you see that there is parallel computing that is 4 times more powerful. And the most interesting thing is that every time we add a qubit, our computing power doubles, that is, we start with two, then move on to four, and with three qubits, our computing power will double again, there will be eight. And four qubits are the parallel processing of sixteen scenarios. This is all very interesting. This is all very new. The computing power of a quantum computer grows exponentially with the number of qubits. They continue to grow exponentially. And let's say you have 50 qubits. In this case, the power of the computer is 2 to 50 degrees. That's many trillions. This is about the same power as in a supercomputer. You can imagine. And if you have 300 qubits, that's only six times more than here. Two to the three hundredth power is more than the number of atoms in the universe. So you can do this kind of parallel computation, with so many bits. ”.

Finally, you see the crescendo. Already a 300-bit quantum computer can replace all our existing and future computers. And what can we say about the 2000-qubit computer created by D-Wave Systems. They miscalculate that they had sold this computer. It was necessary to first catch all the remaining bitcoins, and then just sell the computer time at a very reasonable price. After all, such a computer cannot be completely loaded. For every consumer, such a computer is clean. And the student will be able to solve his problem, and the traveling salesman can get his logistics, and the criminologist entered the intercepted gibberish and received an intelligible answer, and, lost in the maze, will type data on his smartphone, send it to this D-Wave computer, and he is saved. You don't have to walk through the maze and shout at the top of your lungs: get me out of here. The guys missed the firebird. Now, perhaps the cybersecurity forensics firm Temporal Defense Systems led by James Burrell, CTO of TDS, is retrieves bitcoins from the computer. There is less talk about mining on the radio.

Still, John Martinis still has some doubts about the construction of this computer. ”It seems to me that this is working with really big data. Let's see if we can do it. The first part of my lecture is devoted to the fact that we are trying to see if we can build such a computer from 50 qubits. Can we use this quantum computer to calculate that many cases? And really big investments are being made in quantum computers, quantum technologies. Billions of dollars, so let's see if this really works. Let's check it out. ”.

How do you check this? This can be verified only when you build such a computer, and that is not always what the D-Wave scam says. In order not to tire the reader, I will only say that all these transitions of an electron from the ground state to an excited state are carried out by John with the help of a microwave gun (laser), trying to hit this electron in the atom with a photon. Moreover, his photon does not obey any quantum kunstyuk, it moves along a strict trajectory, not like this crazy electron jumping anywhere, duplicating for some reason and the like. So that there is no blunder, this photon hits the squares. You can clarify all this or interpret it differently by listening to the above lecture.

     index         to section    up