"Zu Chongzhi No. 3" is launched! How far are we from using quantum computers?

Recently, Chinese scientists have made a major breakthrough in the field of superconducting quantum computing, successfully building a "Zu Chongzhi No. 3" superconducting quantum computing prototype with 105 qubits, and once again set a new record for quantum computing superiority through the random circuit sampling (RCS) experiment, processing the "quantum random circuit sampling" problem a quadrillion times faster than the fastest supercomputer currently in the world. This achievement was officially published as a cover article in the international authoritative physics journal Physical Review Letters (PRL).

How fast is a quadrillion times faster? What does quantum computing supremacy mean? What does this progress of Zu Chongzhi-3 mean? How far are we from using quantum computers? Let's find out.

(Superconducting quantum computer, copyrighted image from the gallery, reprinting may lead to copyright disputes)

1. How big is the breakthrough of "Zu Chongzhi No. 3"?

Quantum has a superposition state, which means that each quantum bit can encode 2 states at the same time. As the number of quantum bits continues to increase, the number of states that a quantum system can encode will show an exponential explosion.

If we compare a quantum processor to a super brain, then qubits are equivalent to neurons in the brain. The more neurons there are, the more complex and massive the information the brain processes, which means it can solve more difficult problems.

Schematic diagram of the "Zu Chongzhi No. 3" chip.

Image source: University of Science and Technology of China

The Zu Chongzhi III has 105 qubits and can encode 2 to the 105th power states. Compared with its predecessor, the Zu Chongzhi II, which has 66 qubits, the size of the encoding state space of the two is about 500 billion times larger. Of course, this is only the difference in encoding space. When measuring the difference in actual computing power, the fidelity of quantum gate operation and reading must also be considered.

"Fidelity" is a very important concept in quantum computing. Simply put, fidelity is like the "accuracy" of quantum operations, which is used to measure the similarity between actual operations and ideal operations. The higher the fidelity, the closer the actual quantum operation is to the ideal state, and the smaller the error.

The "Zu Chongzhi-3" quantum computer has achieved high fidelity in three key indicators: the parallel single-bit gate fidelity reached 99.90%, the parallel two-bit gate fidelity reached 99.62%, and the parallel read fidelity reached 99.13%. The realization of these high fidelity is like equipping the quantum computer with high-precision "eyes" and "hands", allowing it to execute complex quantum algorithms more accurately.

Therefore, the successful construction of "Zu Chongzhi No. 3" has, on the one hand, greatly improved the upper limit of the computing power of quantum computers, enabling them to handle more complex problems; on the other hand, it has also provided more resources for quantum error correction, and is expected to achieve surface code logic bits with higher code distances, thereby reducing the error rate of quantum computing and promoting quantum computers from laboratories to practical applications.

2. What does the quantum computing superiority that scientists are pursuing actually mean?

When talking about quantum computing and quantum computing superiority, there is one word that cannot be avoided: quantum random circuit sampling (RCS). It is currently an important criterion for measuring the performance of quantum computers.

In the calculation process of a classical computer, the data to be processed is usually input into the system first, and then a series of logic gate operations are performed according to a specific algorithm. After these operations are completed, the processed data obtained is the calculation result. The calculation process of a quantum computer is similar.

Quantum random circuit sampling is to first import quantum information, then run a series of quantum logic gate operations, and finally sample the final quantum state calculation results . The randomness in the quantum random circuit sampling task is mainly reflected in the fact that scientists will randomly select a variety of quantum logic gate operations to obtain a variety of random quantum state calculation results. By performing statistical analysis on these calculation results, the accuracy of the calculation results of quantum computers in various situations can be obtained, and the overall performance of quantum computers can be comprehensively evaluated.

Therefore, it can be said that the better the quantum computer performs in implementing quantum random circuit sampling, the stronger its performance is. When the computing power of a quantum computer on certain specific problems can surpass the strongest classical computer, we call it quantum computing superiority.

The research on "quantum computing superiority" is of milestone significance. It not only verifies the feasibility and potential of the principles of quantum mechanics in the field of computing, provides experimental support for the development of quantum computing theory, demonstrates that quantum computers truly provide computing power that exceeds that of classical computers, promotes the progress of the second quantum revolution, and lays a solid foundation for quantum computers to move towards their ultimate goal - to achieve a fault-tolerant universal quantum computer.

3. How far are we from using quantum computers?

The "Zu Chongzhi No. 3" superconducting quantum computing prototype has achieved the strongest quantum computing superiority in the current superconducting system. Does that mean that we will soon be using quantum computers?

Today, with the soaring complexity of fields such as artificial intelligence and climate simulation, the classical computers we use are unable to meet the exponentially growing computing needs due to the fact that transistors are approaching their physical limits and Moore's Law is gradually failing. Therefore, people place their hopes on quantum computers, hoping that quantum computers can be applied as soon as possible to help us solve the problem of increasing computing power.

However, realizing the practical application of quantum computers is not something that can be achieved overnight.

At present, scientists divide the development of quantum computing into three stages:

Phase 1: Achieve quantum computing supremacy, coherently manipulate more than 50 quantum bits, and have computing power that exceeds the fastest supercomputers for specific problems.

The second stage: to realize a dedicated quantum simulator that can coherently manipulate hundreds to thousands of quantum bits to solve important scientific problems that classical computers are unable to solve, such as quantum chemistry, high-temperature superconductivity mechanisms, and topological states of matter.

The third stage: realizing a universal fault-tolerant quantum computer that can coherently manipulate at least millions of quantum bits with the assistance of quantum error correction, and is used to solve computational problems in the fields of classical cryptography, artificial intelligence, material design, biopharmaceuticals, etc.

To reach the third stage of quantum computer development - universal fault-tolerant quantum computers and realize the full practical application of quantum computing, it is necessary to use quantum error correction algorithms to reduce the error rate of logical bits. Currently, scientists estimate that it will take about 10-15 years to build a quantum computing prototype with thousands of logical bits and a logical error rate controlled at the order of 1e-10.

Reviewer: Zha Chen, postdoctoral fellow at Shanghai Institute of Technology, University of Science and Technology of China, first author of the paper

Planning: Shi Wenhui, Yan Dong

Produced by: Science Popularization China