Entering ChinaSolving a specific problem is a quadrillion times faster than supercomputing! Chinese scientists have successfully developed the prototype of the “Zu Chongzhi No. 3” quantum computing machine
The reporter learned from the University of Science and Technology of China that Pan Jianwei, Zhu Xiaobo, Peng Chengzhi and others of the university have recently successfully built a 105-bit superconducting quantum computing prototype “Zu Chongzhi No. 3”, which can process the quantum random line sampling problem 100 billion times faster than the world’s fastest supercomputer, once again breaking the world record for the superiority of quantum computing in superconducting systems. On March 3, the internationally renowned academic journal Physical Review Letters published the results, and the reviewers said that it “has built the highest level of superconducting quantum computer at present”.
Quadrillion times faster! Chinese scientists have broken another world record
Schematic diagram of the “Zu Chongzhi No. 3” chip. (Photo courtesy of the research team)
Quantum computing is considered to be a key technology for the next generation of information revolution, and the superiority of quantum computing is like a threshold, which verifies the feasibility of quantum computers beyond traditional computers, which is a prerequisite for quantum computing to have application value, and is also a manifestation of a country’s quantum computing research strength.
In 2021, Pan Jianwei’s team successfully built a 66-bit superconducting quantum computing prototype “Zu Chongzhi No. 2”, which solved the quantum random line sampling problem more than 10 million times faster than the world’s fastest supercomputer at that time. After more than three years of concentrated research, their newly developed “Zu Chongzhi No. 3” contains 105 readable bits and 182 coupled bits, and a number of key performance indicators have been greatly improved.
After testing, “Zu Chongzhi No. 3” completed 83-bit 32-layer random line sampling, with the current optimal classical algorithm as the comparison standard, the calculation speed is quadrillion times faster than the current fastest supercomputer, and also millions of times faster than the latest results published by Google in October 2024, which is the strongest quantum computing superiority in the current international superconducting system.
The mainstream view of the international academic community is that the development of quantum computing needs to go through “three steps”: the first step is to realize the superiority of quantum computing; The second step is to develop a quantum simulator that can manipulate hundreds of qubits to solve some problems that supercomputers are not capable of and have great practical value. The third step is to greatly improve the manipulation accuracy, number of integrations and fault tolerance of qubits, and develop a programmable general-purpose quantum computer.
It is reported that the scientific research team of “Zu Chongzhi No. 3” is accelerating exploration in quantum error correction, quantum entanglement, quantum simulation, quantum chemistry and other aspects.
“We are working on surface code error correction with a code pitch of 7, and then expand to 9 and 11 as we make progress, paving the way for large-scale qubit integration and manipulation.” Zhu Xiaobo said.
