Pu Muming Wang Ling
In recent years, quantum computing and quantum computers have attracted a lot of attention in both the scientific and industrial circles. Scientists hope to use the quantum properties of materials to break the Moore's Law of traditional computer miniaturization, and then to establish a new type of quantum computer. The concept of quantum computing was first introduced by the famous physicist Richard · Richard Feynman in 1981. One of the early founders of the field, Mr. Yao Zhizhi, the Turing Award winner, made a core contribution to the establishment of the theoretical foundation of quantum computing in 1993. In 2011, Yao Zhizhi founded the Tsinghua University Quantum Information Center (CQI) to build the latter into a world-class research center for quantum computing. In a recent conversation with the National Science Review (NSR), Yao Zhizhi counted the history of quantum computing and expressed his views on the future development of the field. He believes that the tasks that quantum computers excel at include new material design, drug design, and chemical reaction simulation, but in areas where traditional computers have proven to be highly efficient, they are unlikely to replace them.
NSR: Quantum communication and quantum computing have received widespread media attention. Are the two different concepts?
Yao Zhizhi: Quantum communication and quantum computing are two interrelated, but independent concepts. The technology required for quantum computing is more advanced. The main goal of driving the development of quantum communication is to establish cryptographic guarantees for secure communication. In quantum communication, the signals to be transmitted from one place to another need not be highly accurate. But quantum computing requires high accuracy of the signal. In the past decade or so, large companies such as Google have developed some emerging technologies related to quantum computing. The general view is that the available technology will emerge in the next five or six years. The theoretical basis of quantum computing was established 20 years ago, and the question now is how to implement it.
NSR: Quantum computing has become a hot topic. What is its basic principle?
Yao Zhizhi: The rapid development of semiconductor circuit miniaturization has led to the continuous improvement of the performance of traditional computers. However, this miniaturization has an inherent limit — when the size of the circuit components on the chip is reduced to the nanometer scale, the quantum mechanical effect will dominate and affect the performance of the component. This will be the end of Moore's Law.
For traditional computers, this is an inevitable fate; but scientists have begun to consider whether it is possible to turn harmful quantum phenomena in this case into useful —— construct a quantum mechanical logic using the Schrödinger equation A computer that performs calculations, not a traditional computer that uses Boolean logic for calculations. The idea of quantum computers was first proposed by Feynman in 1981. In principle, he said, one can design a computer that works through quantum mechanical properties, simulates quantum systems and uses quantum equations to solve them. Feynman's philosophy has attracted a lot of attention in the academic field.
Traditional computers use dual-valued Boolean logic (0 and 1) to function through integrated circuits. The calculation is: mapping the input points represented by the bits to a higher level, and by multiple mapping, the output points are obtained to provide the final solution. However, the quantum bits of a quantum computer can represent 1, 0 or any superposition of these two states. The calculation of a quantum computer system is similar to the rotation of a solid; in this analogy, the result of a quantum computer is similar to the reading obtained by measuring the rotation of a solid (the angle of rotation can be any continuous angle). An operation of a conventional computer corresponds to a certain path; an operation of a quantum computer can be performed along multiple computational paths, and ultimately the same target is achieved because the quantum wave function allows multiple states to exist at the same time. This phenomenon is quantum parallelism. Quantum parallel computing is a key reason why quantum computers can be much faster than traditional computers.
NSR: What are the main differences between traditional computers and quantum computers in terms of hardware design?
Yao Zhizhi: The quantum computer is a relatively closed system, and its calculation can be almost instantaneous. Basically, quantum computers behave very “shame”: Once viewed, the calculations are interrupted and stopped. In addition, quantum computers are very complex systems involving multiple frontier technologies. For example, a memory cell of a quantum computer, communication between a plurality of cells, modulation of a quantum bit state, and the like all require a laser. As far as the materials and manufacturing processes of quantum computers are concerned, they not only represent the integration of many advanced technologies in the past three or four decades, but also involve close cooperation among various disciplines.
NSR: Does the uncertainty of quantum phenomena affect the accuracy of quantum computing?
Yao Zhizhi: Yes, but an uncertain answer is not necessarily wrong. In fact, some quantum calculations always give the right answer. And in terms of actual calculations, there are some errors that are acceptable and do not need to be 100% accurate.
NSR: The concept of quantum computers has been around since the early 1980s, but seems to be slow in the next few decades.
Yao Zhizhi: It is true. After Feynman proposed this idea, mainly physicists are conducting in-depth theoretical exploration. Until the early 1990s, after physicists basically clarified the operating mechanism of quantum computers, computer scientists began to enter this field — — I am one of them. In 1994, Peter Shor of Bell Labs designed a quantum computing algorithm for cracking passwords, which led to widespread interest in the computing world. The US government and NASA began to invest in this field. A number of research teams that competed with each other and attempted to make the first actual quantum computer began to emerge.
NSR: Since then, what are the main developments?
Yao Zhizhi: The main work since then is to explore and choose solutions for implementing quantum computers. For the past decade or so, scientists have experimented with materials such as ion traps, superconductors, and diamonds to make quantum computers. Recently, topological insulators have also become one of the alternatives due to their excellent correctable function. But there is still a long way to go before, one of the main difficulties is to maintain the ultra-low temperature of the functional state.
NSR: When do you think the first quantum computer will appear?
Yao Zhizhi: Many people predict that the first quantum computer will appear in the next five or six years, but I think it is not easy to create a quantum computer that can perform reliable calculations at thousands of quantum bits. Large companies such as Google and IBM have invested heavily in quantum computer research and development. In particular, Google recruited John Marinis, the most important expert in the field, and his entire team at the University of California, Santa Barbara.
NSR: At your initiative, Tsinghua University established the Quantum Information Center (CQI) in 2011. What is the goal of this center?
Yao Zhizhi: Our goal is to create a world-class center for quantum information and to train the next generation of scientists in this field. Therefore, our top priority is to recruit high-quality researchers, such as Duan Luming, a professor at the University of Michigan, Fermi, who was recruited. In the past few years, he has done excellent work in our center.
NSR: What are the advantages of using a diamond system?
Yao Zhizhi: The diamond system has two advantages: first, it can operate at room temperature; second, it has a solid crystal structure, and if the system can perform well at several qubit levels, it is possible to expand to a larger scale. In addition to the diamond system, our center is also conducting research on ion traps, superconductors and photonic networks, and is making good progress.
NSR: Quantum computers have excellent performance. Will they replace traditional computers?
Yao Zhizhi: I think traditional computers and quantum computers will coexist because they have their own advantages. Traditional computers have the accuracy and maturity that quantum computers do not yet have. But compared to traditional computers, quantum computers will have an advantage in solving problems involving quantum mechanical effects. For example, in the fields of material design, drug discovery, and physical chemistry, quantum computers will show advantages, and it is difficult to solve these problems using traditional computers.
NSR: The hardware and software of quantum computers are very different from traditional computers. What are the main challenges at the moment?
Yao Zhizhi: Quantum computing is a typical interdisciplinary field that requires close collaboration between scientists and engineers in related fields, especially between quantum physicists and computer scientists. A breakthrough in the algorithm will motivate hardware improvements and vice versa. For example, Professor Peter Shor, who I mentioned above, not only proves that quantum computing can solve the problem of password cracking, but also solves the problem of error correction in quantum computing. It is based on his research that physicists are beginning to be convinced of the viability of quantum computers. The development of the equivalent sub-computer to a certain stage will require a revolution in computer science. Traditional computer data storage, computing systems, and programming languages all need to be redesigned. It is not clear how this will be done, but it is an important research direction. Many leading companies in the IT industry have built a large number of projects to develop quantum software.
The study of quantum computing methods and algorithms is an area of great potential. A number of elegant calculation methods have emerged over the past few decades and are theoretically attractive. I hope to see more quantum computing methods that fit in with reality, such as methods for material design.
NSR: Quantum computers seem to require the common development of science and manufacturing technology?
Yao Zhizhi: That's right. I have already emphasized that in China, the importance of manufacturing quantum computers is far more than just researching quantum computing —— because this will drive the development of related technology industries. This large-scale project will motivate scientists and engineers to create new methods and technologies to solve specific problems. These methods and technologies can make a useful contribution to society in many fields such as industrial development and national security.
There are some immediate concerns when conducting actual experiments. For example, diamond materials suitable for quantum computers rely on foreign imports. As competition becomes more intense, other countries may refuse to sell materials to us. If we don't develop these materials on our own, we will be very limited in the future. In addition, working in this field, it is not possible to publish articles for personal evaluation and promotion in the short term. Unless we change the existing evaluation system, it is difficult to motivate researchers to do this basic research work. These have led to materials and technologies that we still rely on for imports. (Pu Muming: Director of the Institute of Neuroscience, Chinese Academy of Sciences, Executive Editor of NSR; Wang Ling: NSR Contributing Contributor)