China gets ahead in the quantum race
In November 2020 China published their 14th Five-Year Plan (“FYP”), an extensive document that outlines the country’s ambitious plans for the 2021–2025 period. Technology is a core component on this document, with several chapters dedicated to describing how China’s leaders hope to transform the country into an innovation powerhouse: “We will speed up the landscape of quantum computation, quantum communication, integrated circuits, DNA storage and other cutting-edge technologies, strengthening the innovation across information science, life science, materials and other basic disciplines […]” (loosely translated).
The plan is a formal statement to the world and the continuation of a strategy that has started in recent years, with China achieving signifcant breakthroughs in quantum technology, including the first world quantum satellite (Micius), a 2,000 km quantum communication line between Beijing and Shanghai (and an additional one between Hefei and Wuhan), and the recent demonstrations of quantum supremacy over classical methods on specific tasks. New events keep piling up: The country has made significant progress in all areas that are paramount to become a world leader in quantum: along with an official policy comes the capital, business, technology, research, education, conferences, and most important of all, the people pouring their talent and effort into the development of quantum technologies.
One of those talents is Zhang Yiming, a 38-year-old entrepeneur. In April 2021 he topped a list of Chinese business leaders younger than 40 for the third consecutive year, published by the Chinese equivalent of Fortune magazine. Zhang is the founder of ByteDance, and in the nine years since its founding in 2012 he has built his company to a value of almost US$400 billion.
Zhang Yiming (right) CEO of ByteDance, attends the opening ceremony of the 5th World Internet Conference in Wuzhen, Zhejiang Province. Photo: AP
Being the home of the viral video app TikTok, nobody guessed (at least Westerner media didn’t) that ByteDance would have any interest in quantum technologies. Surprisingly, a paper published at arXiv.org on September 17th, 2021 shows that they were the first participating institution in the design of a hybrid quantum and classical algorithm that can efficiently simulate and compute the properties of small molecules. Rather than the common strategies (i.e. VQE), they “integrated an adaptive energy sorting strategy and […] the density matrix embedding theory” to find “a shallower quantum circuit and reduces the problem size“. In other words, it hints the possibility to run heavier calculations with less qubits. This contribution to the field of quantum chemistry marks the official entry of ByteDance into the field of quantum computing.
Another notable startup is Origin Quantum. It was established on September 2017, and it just 4 years has developed subsidiaries in Beijing, Shanghai, Chengdu and Shenzhen City. The company focuses on the full-stack development of quantum computing, and in September of 2020 they released a domestic R&D superconducting quantum computing in the cloud, with the idea to bring the industry and academia closer together. Called Wu Yuan, it has a 6-qubit superconducting quantum processor. The company also has plans for 24-qubit and 64-qubit devices in the future. One of the company owners is Guang-can Guo, professor and doctoral supervisor at the University of Science and Technology of China (USTC), which serves as the company technology base.
USTC and many of these startups are based in the city of Hefei, which efectively has become the country’s hub for quantum technologies, hosting more than 20 quantum-related enterprises with an output value in excess of 430 million yuan (around 66.5 million American dollars) in 2020 alone. In September 18th 2021, the 1st Quantum Industry Conference was held in the city. The conference brought together a group of scientific researchers and industry leaders to discuss the development of the quantum industry.
Optical quantum computer prototype exhibited in the 1st Quantum Industry Conference at Hefei city, Anhui province, China. Photo: Xinhua news
Just within 2020, the group of USTC led by Chaoyang Lu and Yanwei Ping has published 4 papers reporting 4 new quantum supremacy claims. 2 of these 4 papers concerned superconducting qubits, which is the same architecture that Google used in their Sycamore chip when they made their claims in 2019 (and was later refuted by IBM in another paper).
In December of 2020, USTC reported an experiment with 56 superconducting qubits in a chip (in comparison with the 53 that Google racked up in theirs), and they had 20 layers of gates that were applied to these qubits (same number as Google in this case). The fidelity of the circuit, a crucial metric, was about 30% of Google’s. In September 2021 they reported a new sampling-based quantum supremacy experiment using superconducting qubits: this one had 4 more qubits (60) and they did 22–24 layers of gates and a similar gate fidelity to the previous experiment.
The other two papers from USTC concerned Boson sampling, which was the original proposal for sampling based quantum supremacy by Scott Aaronson and his student Alex Arkhipov in 2011. Boson sampling uses instead photons that are generated and sent through a network of beam splitters, and then get measured to see where they end up. In December 2020, USTC reported an experiment with Gaussian-Boson sampling (a variant of Boson sampling) with an average of 50 to 70 detected photons. This beat their own record of 14 photons previously detected, it was the first of this nature and resulted in the first claim of quantum supremacy via optical quantum computing by the Chinese group.
Skeptics have quickly responded over the past 9 months, arguing that probably they could spoof these results using a classical computer. So in June this year, USTC bumped numbers up again with a yet new record of over 100 detected photons in average, and they claimed that they had specifically ruled out classical spoofing strategies, like the ones they were used against their first experiment. The discussion continues to the time of this writting.
One thing that is important to clarify is that there is no claim in any of these experiments (either Google or USTC) of a useful quantum computation being performed, neither the device being truly scalable or their qubits being error-corrected. The claim in all these cases is that some contrived benchmark was much faster than doing the same thing using a digital computer. This is obviously open to challenge, and people have been challenging it indeed by designing classical spoofing algorithms. This becomes effectively a race, and as the spoofing algorithms develop, the experiments are also expanding and improving very rapidly. Today is clear that the USTC is at the forefront of experimentation, and together with Google are the two groups worldwide that are actively in the lead of this sort of experiments.
These are exciting times for quantum technologies. The richest nations in the world want to be at the forefront of innovation, but some are still reluctant to make a serious investment due to the uncertaintly of a potential quantum winter. China has embarked in the race with all its power: The consequences for humanity of complex molecule design, precise and fast machine learning models, or communication channels faster than the speed of light are too enticing to not try. We might not get there anytime soon, but as the Chinese proverb says, it is better to light a candle than to curse the darkness.
