HEFEI, March 4 (Xinhua) -- A team of Chinese scientists on Monday unveiled Zuchongzhi 3.0, a 105-qubit superconducting quantum processor prototype with speed gains in the quadrillions over leading supercomputers for one specific task: quantum random circuit sampling.
The team's groundbreaking result has positioned China at the forefront of quantum supremacy amid rapid global tech progress.
But what does this leap mean, and how could it redefine the future of computing?
BEYOND SUPREMACY
The United States and China are both leading contributors to quantum innovation, advancing the field in parallel and driving breakthroughs that are redrawing computational frontiers.
In 2019, Google's Sycamore declared a quantum advantage over traditional supercomputers in solving a sampling task in 200 seconds, but Chinese scientists challenged that advantage in 2023, cutting the time to 17 seconds using classical chips and algorithms.
Meanwhile, China hit new quantum computing milestones with Zuchongzhi 2.0 and Jiuzhang 3.0. By 2024, Sycamore had again highlighted its quantum supremacy with expanded qubit counts.
But this week, Zuchongzhi 3.0 surpassed Sycamore's latest scores by six orders of magnitude, setting the highest benchmark for a superconducting system ever publicly reported.
Over the past 80 years, computing has reshaped our understanding of the world and daily life, but unlocking its next frontier hinges on overcoming the bottlenecks of processing power as the field approaches the physical limits of Moore's Law.
Quantum tech is broadly considered to be the front line of technology, poised to trigger a transformative breakthrough akin to that of fusion energy.
And scientists racing to embrace this future have laid out some critical milestones.
Zhu Xiaobo, the chief designer of Zuchongzhi 3.0, has outlined a three-stage roadmap for the development of practical quantum computers, the first of which has seen China and the United States showcase capabilities that surpass those of classical supercomputers, such as quantum random sampling carried out primarily through qubit scaling. But these advances remain niche demonstrations with minimal real-world impact.
In the second stage, researchers around the world are aiming to pinpoint a handful of practical quantum applications -- like quantum chemistry and drug discovery, now largely bolstered by supercomputers and AI algorithms -- within five years, translating quantum advantages into tangible productivity gains, according to Zhu.
The third stage will involve achieving universal fault-tolerant quantum computing, which requires suppressing qubit error rates to extreme lows. Given current physical qubit error rates and engineering hurdles, Zhu estimates this milestone remains about 15 years away.
Now, in early development of quantum computing, the front-runners and best technical approaches have not yet been consolidated, meaning "any country that is able to deploy quantum tech first will have a first-mover advantage," according to a report published by the Mercator Institute for China Studies last December.
LOWERING ERRORS
Zuchongzhi 3.0, a superconducting quantum processor with 105 readable qubits, has one of the highest qubit counts of any device that has demonstrated quantum supremacy.
Its world-leading fidelity metrics are its true distinction: its single-qubit gate (99.9 percent), its two-qubit gate (99.62 percent) and its readout (99.13 percent), which have earned peer recognition for what one journal reviewer referred to as "benchmarking a new superconducting quantum computer with state-of-the-art performance."
Zuchongzhi 3.0's ingenious architecture -- integrating frequency-tunable qubits, topological coupling and flip-chip bonding -- has achieved a decoherence performance surpassing that of Sycamore, in a critical advance to maintain quantum states in large-scale systems, according to Zhu.
Quantum coherence time is a measure of how long a quantum system can retain operational integrity, with a longer time enabling the execution of complex algorithms and relatively large-scale computations, Zhu said.
Despite the good performance of the existing Zuchongzhi system, Zhu recognizes that slashing quantum error rates remains a pivotal challenge to the viability of practical quantum computers. "To build a practical quantum computer, we must simultaneously scale qubit counts and reduce error rates."
Last December, Google unveiled a new quantum computer based on the Willow quantum chip that displays excellent error-correction ability.
Willow employs a scalable surface-code quantum error-correction method, which arranges qubits in a checkerboard-like lattice in a strategy that had been verified in China's Zuchongzhi 2.0 in 2022, according to Wu Yulin, a quantum scientist at the University of Science and Technology of China (USTC).
In this recent breakthrough, Google demonstrated surface-code logical qubits with code distances of 7, achieving a significant reduction in logical error rates compared to earlier implementations, according to a paper published in Nature.
Zhu's team is currently conducting surface code error-correction research at an equal level. "Our latest work remains unpublished, but internal benchmarking against Willow's published data has indicated comparable performance metrics at this stage."
Chinese scientists are planning to demonstrate their 7-code-distance error-correction strategy within months. And after they see progress, they will extend that distance to 9 and then 11, paving the way for large-scale qubit integration and control, according to Zhu.
LOOKING AHEAD
Though superconducting qubits often steal headlines when it comes to computing, the manipulation of photons for quantum computing has also seen rapid progress.
China's Jiuzhang quantum system pioneered a sampling problem type with 76 photons in 2020, and scaled that up to 113 photons with Jiuzhang 2.0 in 2021.
In 2022, Canadian startup Xanadu and the U.S. National Institute of Standards and Technology matched this strategy with up to 219 photons. China took the lead again in 2023 with Jiuzhang 3.0 by controlling 255 detected photons.
And China's Jiuzhang 4.0 prototype, designed to integrate over 2,000 photons, is poised to break quantum supremacy benchmarks and has the potential to redefine the competitive landscape of quantum computing.
While quantum scientists explore multiple technical pathways in quantum computing, they are also prioritizing the exploration of its potential to tackle high-impact scientific and engineering challenges.
"We're now in stage two, where scientists are striving to develop quantum simulators for practical applications," said Yao Xingcan, who is based at the Chinese Academy of Sciences' Innovation Academy for Quantum Information and Technology.
Yao was part of a USTC team that developed last year an ultra-cold atomic quantum simulator to address a typical model depicting the behavior of materials in a high-temperature superconductivity state. The team's work outperformed the computational capabilities of any classical computer, demonstrating quantum supremacy in solving the pivotal scientific question.
The discovery of high-temperature superconductors could lead to new possibilities for multiple practical applications, including energy storage, power transmission and new modes of transportation.
The result, according to one peer reviewer, is "an experimental tour de force" that "marks an important step forward for the field" and "could become a notable milestone for modern science and technology, and a major breakthrough."
In a column written for a Chinese audience last December, Nobel laureate Frank Wilczek said that these cutting-edge quantum tools will empower scientists to develop more sophisticated quantum technologies.
In a spiraling cycle of innovation, new aspirations will be ignited and novel realities will be forged, Wilczek said. ■
