A research team at the University of Science and Technology of China (USTC) developed a quantum computing prototype that provides better performance than all existing supercomputers. The superconducting quantum processor Zuchongzhi-3 operates at a speed one million times faster than Google’s Sycamore quantum processor, which had established the benchmark for quantum computing.
Zuchongzhi-3 achieves its high computational speed through its combination of 105 qubits and 182 couplers. Zuchongzhi-3 surpasses Google’s Sycamore quantum processor because it contains 105 qubits while Sycamore only had 67 qubits. Random quantum circuit sampling tasks provided the proof of this breakthrough because they represent a standard evaluation method for quantum system performance. The sharp development of quantum computing technology focuses on developing devices that can handle supercomputing tasks that classical machines need years to finish. The United States and China lead an intense competition to control quantum technology, while China currently holds the lead position.

The Sycamore computer from Google finished its sampling operation within 200 seconds during 2019, while traditional supercomputers would need 10,000 years to complete the same task. USTC successfully completed the quantum computing task in 14 seconds through the utilization of 1,400 Nvidia A100 GPUs during 2023. China progressed through the development of Jiuzhang as their photonic quantum processor and Zuchongzhi-2.1 as a 66-qubit superconducting processor. The University of Science and Technology in China reached quantum supremacy by operating a 255-photon processor in 2023 that delivered performance 10¹? times faster than conventional computers.
Zuchongzhi-3 advances previous developments by adding 105 qubits and achieving 72 microseconds of coherence time. The processor completed an 83-qubit, 32-layer random circuit sampling task, which exceeded the speed of the world’s fastest supercomputers by 15 orders of magnitude. The USTC researchers steer their research towards error correction methods alongside better qubit connection protocols, which will advance practical large-scale quantum computing implementation.