To understand on a superficial level why such a computer is now impossible, it is not necessary to spend years studying quantum physics and the rules of the microworld.
The main thing as an axiom (and otherwise it is impossible) to accept the principle of superposition and entanglement (coherence). The main thing here is not to confuse the sections of physics and discover quantum mechanics. You can force a textbook, you can Wikipedia, or you can open an old article, which at the same time will tell you about the principles of quantum computing. Very accessible - on the fingers.
compress.ru/article.aspx?id=17653
or in pictures
Promotional video:
But you better take a look at the article on the link above too.
First, you need to honestly and bluntly say that no quantum computer will help your favorite toy run faster under Windows. At this half of those interested should fall off. We will miss you. Others can, if interested, listen to the author of the quantum computer kipish David Deutsch (in Russian).
scisne.net/a-526?pg=11
As you can see, even he was aware of the main problem. Decay of qubit bonds. Decoherence.
Possible states of a memory cell. On the left is the current trigger, on the right is a steep but highly mythical qubit.
A lonely qubit is essentially a useless thing, despite all the fantastic laws of quantum physics. Only bundles of qubits that are in a state of quantum entanglement, that is, that affect each other, are important for high-speed computations. The more elements in a bundle, the better. Building a quantum computer in the form in which we represent it is currently scientifically (!) Impossible. At least of that power, where the advantages of an infinite enumeration of states of quantum systems will be clearly visible to the naked eye.
This requires a coherent (bound, entangled, etc.) state, at least a couple of hundred quanta (min> 103), and preferably a couple of thousand. At the moment, a bunch of 8-10 is fully functional.
Decoherence makes long quantum computations impossible. At first, it was believed that it arises from external pressures on the "system" and attributed the problem to the category of engineering. However, experiments with different technologies for constructing qubits (quantum dots, superconducting elements, atoms and ions in traps, etc.) have shown that it is of a fundamental nature and originates in the “collapse of the wave function”.
The very effect of the observer. You can start studying this issue here, where some Cambridge and not only employees quarreled in open correspondence.
www.lightbluetouchpaper.org/2013/02/01/har…
In fact, there is no consensus. Someone thinks that decoherence can be avoided, and then the dream of Quantum Supremacy is realizable, someone is skeptical, but even the biggest problem of quantum computers is not the only obstacle to creating a miracle.
All current attempts to assemble "something" called a "quantum computer" do not apply to conventional computers at all. All research is aimed only at the application of Shor's algorithm and breaking RSA encryption.
Evolutionary modeling of protein compounds for medicine is sometimes mentioned to cover their tracks, but this, as they say, is not accurate. They seemed to have modeled the molecule, but they immediately made a reservation that due to computational errors, it is not very similar to the real one.
www.sciencemag.org/news/2017/09/quantum-com…
In principle, all calculations based on qubits are very accurately characterized by the phrase “but this is not accurate.” The very possibility of achieving an imputed result from a bundle of quanta is based on the cyclical execution of a function and achieving an approximate result by multiple (very multiple) repetitions.
The laws of Moore, evolution, man-hours and Einsteins per square meter are inapplicable here, because with each quantum attached to the chain, the effect of errors on the result grows exponentially. This means that the number of searches for a probabilistic, close to correct, answer begins to tend to infinity. And if not to her, then certainly to the capabilities of the quantum computer itself - for sure. And all Supremacy will go to steam.
Let's add here a limited set of formulas available for calculations. There are very few of them and the most striking, so far, is Shor's algorithm. Cryptocurrency dealers are especially worried about this. Afraid of the death of encryption protocols. After all, it will lead to the instant death of all current quasi-currencies.
It really is. But it will be such an insignificant event in comparison with the worldwide effect of the fall of RSA that it will not be noticed. If there is no breakthrough in theoretical or practical quantum physics in the near future, then all efforts will be limited to the decomposition into prime factors and the destruction of RSA.
Are there any breakthrough tendencies? Have fun with making decoherence-protected quantum memory. 100 milliseconds and 22% read is a record. As the beginning of the path, it is probably good, as an application for the nearest implementation, very sluggish.
www.nature.com/articles/s41566-017-0050-y
They are glad that the causal relationships in quantum computing do not correspond much to those we are used to. Given the roughness of all quantum calculations, the hope of getting the correct result is still far away.
www.nature.com/articles/s41567-017-0008-5
With grief, we kept 20 qubits together in half, 50 are on the way (but this is not certain), but the results are not encouraging. Now they propose to operate not only with the number of qubits, but also with the depth of calculations. How many logical operations will the "striped swimsuit group" perform until the tigers have dinner time decoherence?
www.nature.com/news/race-for-quantum-supre…
Best of all, as long as models of quantum computers are obtained, solved by standard computing systems.
www.nature.com/articles/nature24622
So, the "quantum computer" is the same "spherical horse in a vacuum" that we all love so much.
PS
Similar to sub-buffers.
No, I haven't forgotten. It's just, somehow embarrassing to mention him. Or it is not clear from which side. There is no scheme, which we understand as a "quantum computer", but there is a simulation of such systems by quantum annealing. Secondly, its entire mega-cubicity (today 2K) consists of beams of 4 (8) pieces, which means that there is no scaling for which everyone is chasing. They began to suspect something long ago.
physicsworld.com/cws/article/news/2014/jun/…
Then, nevertheless, they adapted it to some kind of optimization of calculations. And this is the worst thing! So "it" works. Albeit crooked, but better than the standard one, which is "no way."
www.nature.com/news/d-wave-upgrade-how-sci…
From the height of knowledge of the world "from the basement", the story of D-Wave seems so mysterious to me that it casts doubt on other works on creating quantum computers and leads them towards conspiracy theory. Nevertheless, I still expect a breakthrough and will carefully follow behind such developments in order to go gold in time.