A research team at the University of Science and Technology of China (USTC) has unveiled a prototype of a superconducting quantum processor, “Zuchongzhi-3,” equipped with 105 qubits. It has achieved a significant breakthrough in accelerating quantum calculations, solving a random sampling problem from quantum circuits a thousand trillion times faster than the fastest supercomputer, and a million times faster than the latest processor presented by Google last year.

This superconducting quantum computer prototype demonstrates high operational accuracy in executing quantum operations, exceeding 99%, and the qubit state readout accuracy after calculation reached 99.18%.

The Zuchongzhi 3.0 processor represents a major upgrade from its predecessor, Zuchongzhi 2.0, which was developed in 2021 and equipped with 66 qubits. With improvements in electronic circuits and a noticeable increase in both the quantity and quality of qubits and their connectivity, it now includes 105 qubits, arranged in 15 rows and 7 columns, forming a two-dimensional rectangular grid.

It was tested on a random circuit sampling task, a standard method for evaluating the performance of quantum processors. The most powerful traditional supercomputer, Frontier, would require about 6.4 billion years to complete it, while the new quantum processor solved it in a few hundred seconds, surpassing the fastest previous quantum processor by about a million times. This was Google’s 67-qubit superconducting processor, Sycamore, presented in October 2024.

It is worth noting that this task has no practical application and serves as a benchmark to demonstrate the computational advantage of quantum processors over classical supercomputers. It involves running complex quantum circuits with random configurations and sampling from the output distribution.

The team published their research results in the journal Physical Review Letters, and they are working on further developing the research to create a processor with a larger number of qubits and higher efficiency, aiming to achieve a quantum computer capable of solving practical problems with high efficiency.