The first quantum computer in the world. A quantum computer in russia, myth or reality

A group of physicists from the USA and Russia has created and successfully tested a programmable quantum computer based on 51 qubits. This was reported in a press release from the Russian Quantum Center, which was received by Indicator.Ru.

Many scientific groups are now trying to create a universal quantum computer, and many governments and corporations are investing in these projects. The computational elements of such computers - qubits - operate on the basis of quantum objects: ions, cooled atoms or photons, capable of being in a superposition of several states. This allows quantum computers to simultaneously, in one clock cycle, do many calculations at once. Quantum computers will be able to cope with tasks that would take classical computers billions of years to solve.

The capabilities of quantum computers depend on the number of qubits. Already several tens of qubits can give such a gain in computing power, which is unattainable for classic computers. Today, the quantum laboratory of Google corporation under the direction of John Martinis is planning experiments on a computer with 49 qubits, IBM is already experimenting with a 17-qubit device. The creation of a 51-qubit computer is a giant step forward in this area.

A group of scientists from Harvard University and the Massachusetts Institute of Technology led by Mikhail Lukin, a physics professor at Harvard and co-founder of the Russian Quantum Center, used qubits based on cold atoms, which were held by optical "tweezers" - specially organized laser beams. Most modern quantum computers are based on superconducting qubits based on Josephson contacts.

Lukin and his colleagues managed to solve with the help of their quantum computer the problem of modeling the behavior of quantum systems from many particles, which was practically unsolvable with the help of classical computers. Moreover, as a result, they were able to predict several previously unknown effects, which were then tested using conventional computers.

In the near future, scientists intend to continue experiments with a quantum computer. Perhaps they will try to use this system to test quantum optimization algorithms that can outperform existing computing machines.

According to Lukin, who made a presentation at IV International conference on quantum technologies in Moscow (ICQT-2017) July 14, an article with the results of the work has been accepted for publication and will appear on the arXiv preprint server on Sunday. On the evening of July 14, Lukin will take part in an open discussion at the ICQT conference, which will take place after a public lecture by John Martinis.

On Friday morning, July 14, at the International Conference on Quantum Technologies, Mikhail Lukin, a co-founder of the Russian Quantum Center and a professor at Harvard University, spoke about the creation by his research group of a fully programmable 51-qubit quantum computer. At first glance, this result can be called a sudden breakthrough in this area - such giants as Google and IBM are just approaching the 50-qubit mark in a quantum computer. Just yesterday on the arXiv.org preprint server appeared detailed description experiment. Editorial staff N + 1 decided to figure out what happened and what to expect from the new quantum computer.

Briefly about quantum computers - universal and non-universal

What is a 51-qubit computer like?

Let's deal with the system created by physicists in new job... The role of qubits in it is played by cold rubidium atoms captured in an optical trap. The trap itself is an array of 101 optical tweezers (focused laser beam). The atom is held by tweezers in equilibrium position due to the gradient electric field- it is attracted to the area with the maximum electric field strength, which is located at the focal point of the tweezers. Since all the tweezers are lined up, all the atom-qubits of the computer are also lined up.

"Zero" for each of the rubidium atoms is its basic, unexcited state. “One” is a specially prepared Rydberg state. This is such an excited state in which the outer electron of rubidium is very far from the nucleus (at the 50th, 100th, 1000th orbital), but still remains associated with it. Due to the large radius, Rydberg atoms begin to interact (repel) at much greater distances than ordinary ones. This repulsion makes it possible to transform a row of 51 rubidium atoms into a chain of strongly interacting particles.

A separate laser system is used to control the states of qubits, capable of exciting them to the Rydberg state. The main and most important feature of the new computer is the ability to directly address each of the 51 qubits. There are also more complex ensembles of atoms in which entangled quantum states are observed (recently we have about 16 million atoms entangled by interaction with one photon), and quantum simulations have also been performed on more than a hundred cold atoms. But in all these cases, scientists did not have the ability to accurately control the system. This is why the new system is called a fully programmable quantum computer.

Every computation on a quantum computer is, in a sense, a simulation of a real quantum system. The bulk of the new work is devoted to modeling a well-known quantum system - the Ising model. It describes a chain (in this case) of particles with nonzero spins (magnetic moments) interacting with their neighbors. The Ising model is often used to describe magnetism and magnetic transitions in solids.

The experiment was structured as follows. The particles were first cooled and captured in optical tweezers. This is a probabilistic process, so at first the array of particles was chaotic. Then, using a sequence of measurements and corrections, a defect-free array of more than 50 cold atoms in the ground unexcited state was created. At the next stage, the optical tweezers were turned off and at the same time turned on the system that excited the atoms to the Rydberg state. For some time, the system evolved under the influence of van der Waals forces - atoms occupied the most "convenient" positions for them, after which the tweezers were turned on again and the result of evolution was studied.

Physicists observed different evolutionary results depending on how close the cold atoms were before the exciting pulse. This is due to the fact that Rydberg atoms are capable of suppressing the excitation of neighbors to Rydberg states (due to strong repulsion). Scientists have observed systems in which atoms after evolution turned out to be ordered so that between each pair of neighboring Rydberg atoms there was strictly one, strictly two, or strictly three ordinary atoms.

Interestingly, the formation of highly ordered structures after free evolution occurred with a very high probability - even in the case of an array of 51 cold atoms.

To see how the evolutionary process takes place, scientists turned on tweezers and "photographed" the system at different points in time. It turned out that in some cases the evolution to a state of equilibrium proceeded very slowly: the system oscillated for a long time between several states. This result can be confirmed by rough classical modeling involving the interaction between neighboring and following neighboring atoms in the analysis.

Is it helpful?

This is one of those cases where quantum modeling predicts a real new effect. It is worth noting that it is impossible to accurately simulate a system of 51 cold atoms using a classical computer. To just describe all its possible states, you will need 2 51 bits of RAM (about a petabyte). This effect was confirmed only by rough modeling on a classical computer.

It is interesting that exactly the opposite situation arises in quantum chemical calculations - classical computers give only an approximate estimate of the properties for complex systems, spending huge computational resources on this. At the same time, direct analysis of these, of course, quantum systems gives an accurate result.

And what else is it useful for?

At the end of the preprint, the authors traditionally provide a list of areas in which new development can be useful. Some of them can be listed: the creation of superpositions consisting of a large number of particles, the study of topological states in spin systems. Physicists point out separately that the algorithm is well suited for solving problems of optimizing systems, the sizes of which obviously exceed the reach of conventional computers. These tasks include simulation of chemical reactions and training.

The system created by Mikhail Lukin and his colleagues works now as a quantum simulator - it simulates systems similar to itself. However, it is worth noting that physicists have already managed to create logical CNOT-valves used to create entanglement on individual pairs of Rydberg atoms. Therefore, we can say that some of the simplest algorithms can be implemented in the new system (for example, Deutsch's algorithm, or Shor's algorithm for very small numbers). However, at this stage, these algorithms will not be useful.

Mikhail Lukin (left) and John Martinis (right) - head of the 49-qubit quantum computer team at Google

Russian quantum center

In a sense, the new device is already capable of solving problems inaccessible to classical computers - it cannot be accurately simulated by conventional computers. But it is too early to talk about useful quantum superiority, which is already useful in applied problems. Many scientists point out that the race for quantum supremacy now does not carry anything useful from the point of view of applied computational problems.

It is worth noting that experiments with atoms in optical lattices already several years ago surpassed the limits of accurate modeling by classical computers. They use dozens of interconnected particles. For example, with their help, quantum cooperative phenomena related to superfluidity and superconductivity. Is this quantum superiority?

Vladimir Korolev

Within the framework of the International Conference on Quantum Technologies ICQT-2017, which is held under the auspices of the RQC in Moscow, Professor of Harvard University, co-founder of the Russian Quantum Center (RQC) Mikhail Lukin said that a group of Russian and American scientists working at Harvard under his leadership had created and tested the world's first quantum computer, consisting of 51 qubits - the most complex computing system in existence today.

Quantum computers - special computing devices, whose power grows exponentially due to the use of the laws of quantum mechanics in their work, consist of qubits - memory cells and at the same time primitive computing modules that store a range of values ​​between zero and one.

Such devices are developed using the classical or adiabatic method. Proponents of the first are trying to create a universal quantum computer, in which qubits would obey the rules by which ordinary digital devices work. Working with it is similar to how engineers and programmers control computers. An adiabatic computer is easier to create, but it is closer in its principles of operation to analog computers of the early 20th century, rather than to traditional digital devices.

In 2016, several teams of scientists and engineers from the United States, Australia and several European countries announced that they would create such a machine in the near future. So, the team of John Martinis from Google has developed an unusual "hybrid" version of the universal quantum computer, which combines elements of analog and digital approaches to calculations.

Physicist Lukin and his colleagues at RCC and Harvard bypassed Martinis' group, which is now working on the creation of a 22-qubit computing machine, using exotic "cold atoms" rather than superconductors, as scientists from Google.

For example, Lukin's group discovered that a set of atoms, held inside special laser "cells" and cooled to ultra-low temperatures, can be used as qubits of a quantum computer, maintaining stability under a fairly wide range of conditions. This allowed physicists to create the largest quantum computing device to date with 51 qubits.

Using a set of such qubits, several physical problems were solved, which are extremely difficult to simulate with the help of "classical" supercomputers. Scientists were able to calculate how it behaves big cloud particles connected with each other, and detect previously unknown effects arising inside it. It turned out that with the damping of excitation in the system, some types of oscillations can remain and be held virtually indefinitely, which scientists did not suspect before.

For this, a special algorithm was developed that allows similar calculations in a very rough form on conventional computers... The results were broadly consistent, confirming that the 51-qubit system of Harvard scientists works in practice.

The team of scientists intends to continue experiments with a quantum computer. According to Lukin, they will try to run Shor's famous quantum algorithm on it, which allows them to crack most existing systems encryption based on the RSA algorithm. The results of a quantum computer have already been described in one of the peer-reviewed scientific journals.

During the International Quantum Conference in Moscow, Russian scientist Mikhail Lukin presented the most powerful 51-qubit quantum computer to date. The number 51 was not chosen by chance: Google has been working on a 49-qubit quantum computer for a long time, and therefore it was a matter of principle for Lukin, as a gambling scientist, to bypass a competitor.

“A quantum computer is functioning, it is much more terrible than an atomic bomb,” says Sergei Belousov, co-founder of the Russian Quantum Center. - He (Mikhail Lukin) made a system with the most qubits. Just in case. On this moment I think that's more than twice as many qubits as anyone else. And he made 51 qubits on purpose, not 49. Because Google kept saying that they would make 49. ”

However, Lukin himself and the head of the quantum laboratory at Google John Martinez do not consider themselves competitors or rivals. Scientists are convinced that nature is their main rival, and the main goal is the development of technologies and their implementation to advance humanity to a new stage of development.

“It’s wrong to think of it as a race,” says John Martinez. - We have a real race with nature. Because it's really hard to build a quantum computer. And it's just exciting that someone managed to create a system with such big amount qubits. So far, 22 qubits is the maximum we could do. Even though we used all our magic and professionalism. "

The qubits themselves, in the number of which scientists “compete” so furiously, are a computational unit that simultaneously represents both zero and one, while the usual bit is either one or the other. Modern supercomputers build sequences, and quantum computers, in turn, perform calculations in parallel, in an instant. Thanks to this approach, computations that today's supercomputers will take thousands of years to accomplish can be done instantly by a quantum computer.

“This is one of the largest quantum systems ever created,” says Mikhail Lukin, a professor at Harvard University and co-founder of the Russian Quantum Center. “We are entering a regime where classical computers cannot cope with computations. We make small discoveries, saw new effects that were not expected theoretically, which we can now, we are trying to understand, but do not fully understand. "

So far, even the creators of the most powerful quantum computers cannot say for sure why mankind will need such powerful computers. Perhaps, with their help, fundamentally new materials will be developed. New discoveries can be made in the field of physics or chemistry. Or perhaps quantum computers will finally help us fully understand the nature of the human brain and consciousness.

“When a scientific discovery is made, its creators do not represent the full power that it will bring,” says Ruslan Yunusov, director of the Russian Quantum Center. - Here is an example of a transistor. When the transistor was invented, no one imagined that computers would be built on this transistor. And when they built computers, no one imagined how much life would change. "

Russian and American scientists from Harvard University, working in Mikhail Lukin's group, have created a 51-qubit quantum computer, the most powerful in the world today. The co-founder of the Russian Quantum Center (RQC), Professor Lukin, announced this in his report at the International Conference on Quantum Technologies (ICQT-2017), which was held in Moscow in July under the auspices of the RQC.

Unlike classical digital computers, in which the memory is built on the principle binary code(0 or 1, "yes" or "no"), quantum computers build on the basis of qubits - quantum bits. They also admit two states (0 and 1), but due to their quantum properties, the qubit additionally also admits superposition states, that is, conventionally speaking, there are still a lot of intermediate states between two ground states described by complex (imaginary) numbers. It is clear that under such conditions the power and speed of a quantum computer are several orders of magnitude higher.

The very idea of ​​using quantum computing to solve purely math problems proposed back in 1980 by Yuri Manin from the Steklov Institute, and a year later the principle of building a quantum computer was formulated by Richard Feynman. But it took decades before technologies appeared that could put their ideas into practice.

The main problem was to create stable working qubits. Lukin's group did not use superconductors for them, but the so-called cold atoms, which are held inside laser traps at ultra-low temperatures. This allowed physicists to create the world's largest quantum computing machine of 51 qubits and bypass their colleagues the group of Christopher Monroe at the University of Maryland (5-qubit device) and John Martinis's group of Google (22-qubit device).

Figuratively speaking, during the construction of a qubit computer, physicists returned from digital to analog devices the first half of the last century. Now their task is to move to digital on a new, quantum level. Using a set of qubits based on "cold atoms", Lukin's team has already been able to solve several particular physical problems that are extremely difficult to simulate with classical computers.

In the near future, scientists intend to continue experiments with a quantum computer. In addition to solving purely scientific problems in the field of quantum mechanics, Professor Lukin does not exclude that his team will try to implement Shor's famous quantum algorithm on it, before which the existing encryption systems are powerless. But there are many other practical areas where a new generation of computers could revolutionize. For example, hydrometeorology, where now there is clearly not enough power of existing computing devices to improve the accuracy of weather forecasts.

Quantum computers are making their first steps, but the time is not far off when they will become as commonplace as today's PCs.