We are facing a historical time when Quantum Information Technology (QIT) has recently advanced to a tipping point where small-scale prototypes of programmable quantum computers have been demonstrated in academic and industrial labs worldwide. Major industrialized countries and multinational corporations continue to drive investment, research, and development in an attempt to retain technological leadership and dominance in this rising field. Concurrently, private funding is pouring into dozens of start-up companies, hoping to translate quantum-enabled technologies into commercialization. Yet, while the progress and prospect of QIT are promising, there are currently two challenges in the forthcoming future – how to realize large-scale quantum processors, and how to harness the power of quantum computing to design effective quantum information technologies with desirable performance guarantees. 

Prof. Hao-Chung Cheng’s research focuses on the latter problem of theoretical performance analysis on quantum information-processing systems. Specifically, our studies characterize the trade-offs between the required resource, the system size, and the system performance of certain quantum communication and quantum learning tasks. Hopefully, the results can serve as a guideline for designing the next-generation quantum information-processing technologies.

Apart from basic exploration of quantum mechanics in the 1930s and 1940s, the newly developed Quantum Information Science in the late 1970 exploited the framework of quantum mechanics to reformulate a fundamental unit of information, the bit, and applied quantum gates to the so-called quantum bits (qubits) for logic operations and processing. This yielded a series of new emerging fields such as quantum computing, quantum communication, quantum cryptography, quantum simulation, quantum metrology, and quantum machine learning, paving the way for next-generation information technologies.

Compared with classical information systems, a quantum information system has unique features –superposition and quantum entanglement. Based on these features, quantum advantages have been demonstrated in various fields of information engineering. For example, by measuring on a uniform superposition state of 0 and 1, one can generate true uniform random numbers. This procedure is the so-called Quantum Random Number Generator, which processes higher entropy and yields high-level security applications in cryptography compared with today’s Pseudo Random Number Generator. Computation is a core technology in our daily lives and information industries as well. Quantum algorithms based on Quantum Parallelism can solve specific computational tasks such as unstructured searching, integer factorizations, and linear equation solving, with much lower computational complexity than the best classical algorithms heretofore. The applications of quantum computing include quantitative finance, machine learning, and related artificial-intelligence-based services. One can potentially apply a quantum computer with large amounts of qubits to simulate many-body systems in biology and chemistry, such as Nitrogen fixation and new drug findings. Entanglement-assisted communication can increase the channel capacity and achieve secure communications. Quantum communication can be used for classical key distributions with provable information-theoretical security. As such, quantum information technologies have been demonstrating numerous advantages and have shown potential in some fields. Ergo, information and computer scientists worldwide have been urging to explore more aspects of technologies where there are significant quantum advantages. 

Unfortunately, quantum information technologies are not able to change our daily lives immediately. It is because this field faces two enormous challenges from the hardware and software aspects. Regarding hardware, candidates of implementing a gate-based programmable quantum processor unit (QPU) include superconducting, semiconductors, quantum photonics, and ion traps. The state-of-the-art superconducting QPU only has under 100 physical qubits available. It cannot solve integer factorization problems that current classical supercomputers could not solve, as it would require up to 2,000 logical qubits for implementing Shor’s algorithm. On top of that, most quantum algorithms require high-fidelity quantum bits and gates to work correctly. Hence, one logical qubit would need hundreds of or even one thousand physical qubits for error correction. That is, scientists have to build a QPU with at least one million physical qubits to truly achieve universal quantum computation and demonstrate quantum advantages. Speaking of the software, the transition from Quantum Information Science to Quantum Information Technology has to go through several phases such as modularity, systemization, and standardization. Quantum algorithm or protocol designers have to provide performance analysis according to the underlying environmental noises (for example, the infidelity of quantum gates or the noise in quantum channels), to properly configure and control the quantum system. Nonetheless, due to the architectural complications and the inherent non-commutativity of quantum mechanics, it is extremely difficult to conduct theoretical analysis. Furthermore, existing analytical techniques in classical information and computer science will not apply. Consequently, software developers and theorists have to research novel analytical methods to develop next-generation quantum information systems. In summary, building quantum information technologies is a ten-year-long project. It requires multi-disciplinary experts such as mathematicians, physicists, computer scientists, and electrical engineers to collaborate and conquer the forthcoming challenges and obstacles.  

The academic research of Professor Hao-Chung Cheng and his team focuses on the latter challenge – system design and theoretical performance analysis of quantum information systems. The core idea of this research is studying a fundamental problem: how many resources do we invest and how large a scale of system do we need to achieve the desired system performance when implementing a quantum information-processing task? The trade-off between the system size and the system performance is also under investigation. Prof. Cheng’s research includes the following foundation protocols of quantum information systems: (i) the error analysis, communication rate, and coding size in communicating information through a quantum channel or quantum network; (ii) the error analysis, compression rate, and coding size in data compression with quantum side information; (iii) proposing learning algorithms for quantum state tomography and certain quantum circuits, and deriving upper bounds on the sample complexity. The research outcomes mentioned above will be beneficial to developing the next-generation quantum information processing protocols and can serve as a system design guideline. In the near-term future, Prof. Cheng will apply the established analytical techniques to explore various and substantial quantum information processing systems.

Although recently there has been significant progress in building the QPU and studying its theories worldwide, the development of quantum information technologies in Taiwan is still nascent. Skilled experts in research and development are still in high demand. To face the stringent competition of quantum technologies around the world, we require industry, government, universities, and academia to work together to bring the quantum information technologies to fruition.