The article is a guess about the properties of a quantum computer, announced by Google. The company claims that "a quantum computer was able to accomplish what modern computers would have taken thousands of years." The magazine asks the question: does the Google team understand the car they created? And will she not "optimize" the person?
In a recent article from Google's Quantum Computing Lab, it is announced that the company has achieved quantum supremacy. This is just talk, but what does it all mean?
In 2012, I coined the term “quantum superiority” for the point where quantum computers can do what classical computers cannot - and regardless of the utility of the tasks. With this new term, I wanted to emphasize what a historically important time we live in. It is an honor to live in the era of information technology development based on the principles of quantum physics.
The term "quantum superiority" - and indeed the phenomenon itself - has caused a lot of controversy. And for two reasons at once. First, the word "superiority" sounds bad from a political point of view - it evokes disgusting associations with "white superiority." Secondly, this term only exacerbates the general hype around quantum technologies, and there is already too much of it. I still foresaw the second, but I completely missed the first. Be that as it may, the term stuck, and now the quantum team at Google has taken up it with enthusiasm.
I went over a couple more options in my head, but I rejected all of them, deciding that "quantum superiority" reflected the situation in the best way. I also pondered over "quantum advantage" - and this term also came into use. But for my taste, "superiority" still sounds more accurate and convincing. On the race, they talk about an advantage even when one horse is ahead of the other by less than a body. The speed of a quantum computer for certain tasks is several times higher than the classical one. In theory, anyway.
A recent Google article illustrates this clearly. Using a device with 53 qubits (quantum analogs of the bits of a classical computer), they were able to perform quantum computations in minutes that would take thousands of years for today's most powerful supercomputers. If this is true, then this is an outstanding achievement in experimental physics and evidence of an unprecedented development in quantum hardware. I sincerely congratulate all the participants in the experiment.
However, there is a catch here. The Google team admits that the problem that their machine solved with astonishing speed was carefully chosen to demonstrate the superiority of a quantum computer. There is no practical sense in solving it. In short, the quantum computer executed a randomly chosen sequence of instructions, after which all qubits were measured, receiving a bit string at the output. This is a quantum computation of a very small structure. Yes, such a task is extremely difficult for a classical computer, but the answer is not very meaningful either.
And yet the result is outstanding. By checking that the output of their quantum computer matches the output of a classic supercomputer (well, if the operation doesn't take thousands of years, of course), the team confirmed that they understand their device and that it works as expected. Now that we have figured out that the hardware is working, it can be loaded with more useful tasks.
Promotional video:
Why is it important to check the performance of the equipment? Because it is very difficult to precisely control a quantum computer. In a sense, a simple consideration of a quantum system inevitably violates it - the famous Heisenberg uncertainty principle in action. Therefore, if we want to use such a system for storing and reliable processing of information, we need to keep it almost ideally isolated from the outside world. At the same time, we need the qubits to interact with each other - we want to process data. In addition, we have to control the system from the outside and ultimately measure the qubits to find out the result of our calculations. It is extremely difficult to create a quantum system that meets all of these criteria. To achieve the current results, it took many years of progress in the field of materials,production, development and control.
Google's achievement is a milestone in the development of applied quantum computers. I thought that the coming era needed a separate name - and came up with the "noisy intermediate level quantum" or NISQ. Rhymes with risk. “Intermediate Level” is about the size of quantum computers that are becoming more affordable. At this rate, they will soon be able to perform tasks that are too heavy for today's supercomputers. "Noisy" emphasizes imperfect control and the resulting failures and errors that accumulate over time. We cannot perform a long calculation yet - the correct answer is unlikely to shine for us.
The Google team has proven that it is possible to build a quantum machine large and accurate enough to solve previously impossible tasks. Let's consider this the beginning of a new era - the era of noisy intermediate-level quanta or NISQ.
What will happen next? Of course, Google and other hardware makers expect to find practical uses for their quantum devices. A more powerful quantum computer will help scientists develop new materials and chemical compounds or improve tools for machine learning, but a noisy quantum computer with a few hundred qubits will do little good. However, we have some practical ideas for NISQ computers that we are going to try out. By doing this, we will achieve better optimization methods and more accurate physics simulations - although, in truth, we are not quite sure of success. But it will still be interesting to play with NISQ technology to see what it can do. I am countingthat quantum computers will sooner or later transform our society - even if this is a matter of the future.
In a 2012 article where I coined the term “quantum supremacy,” I reasoned, “What is it like to manage large-scale quantum systems? Just "very difficult" or "ridiculously difficult"? If the former, then after a couple of decades of hard work we may well succeed. If the latter, then it will take several centuries, and even then it is not a fact that it will work. " Recent advances by the Google team lead us to believe that we are dealing with the first case - just "very difficult." If so, an abundance of quantum technologies will appear in the coming decades.
James O'Brien