Zuchongzhi-3
Zuchongzhi-3 is a 105-qubit superconducting quantum computer that can perform calculations 10¹⁵ times faster than the most potent supercomputer and one million times faster than Google’s most recent quantum results, according to research from a team at the University of Science and Technology of China (USTC).
Quantum computing has advanced significantly with Zuchongzhi-3. Leading the research team that announced the discovery was Jianwei Pan. Chengzhi Peng, Xiaobo Zhu, and other team members published their results in Physical Review Letters.
When a quantum computer surpasses even the most potent classical computers in calculation speed, this is known as quantum supremacy. For instance, the 53-qubit Sycamore processor from Google in 2019 finished a particular sampling action in 200 seconds, but a classical supercomputer would take 10,000 years to assemble.
Researchers at USTC, however, discovered in 2023 that an enhanced classical algorithm could use more than 1,400 A100 GPUs to do the same operation in 14 seconds. Google’s previous assertion of quantum supremacy is called into question as the operation can now be finished in 1.6 seconds with the Frontier supercomputer’s increased RAM.
Zuchongzhi-3 Technical specifications and performance
Zuchongzhi-3, an improvement on the 66-qubit Zuchongzhi-2, performs better than earlier models in several important areas. The system features a 72 microsecond coherence time, a 99.90% single-qubit gate fidelity, a 99.62% two-qubit gate fidelity, and a 99.13% readout fidelity. More intricate and precise quantum operations are made possible by the advancements.
The research team outperformed the most potent classical supercomputer by 15 orders of magnitude in processing speeds when they completed an 83-qubit, 32-layer random circuit sampling assignment. Additionally, Zuchongzhi-3 performed six orders of magnitude better than Google’s most recent quantum computing accomplishments, establishing a new standard for superconducting quantum systems.
Techniques and Design of Experiments
With 105 transmon qubits, the team at the University of Science and Technology of China created Zuchongzhi 3.0, which is a huge increase over Zuchongzhi 2.0. A 15-by-7 qubit rectangular lattice is used in the processor’s design, and 182 couplers are integrated to improve connectivity. For their experiment, the researchers chose 83 qubits and optimised them for increased stability and lower error rates.
The group carried out a large-scale random circuit sampling experiment to assess performance. In this procedure, a series of randomly selected quantum operations are carried out, and the output of the system is then measured. The exponential complexity of quantum states makes it difficult for classical supercomputers to reproduce this process.
The previous random circuit sampling record set by Google utilised a 67-qubit device operating at 32 cycles. This complexity was raised by the Chinese team, which used 83 qubits and 32 cycles of depth in their experiment. Compared to Google’s 67-qubit experiment, the projected classical computational cost to simulate this system was six orders of magnitude higher.
Restrictions and Difficulties
The paper notes that scaling quantum computing continues to present difficulties despite its performance improvements. Multi-qubit operation errors continue to be a challenge, especially as circuit complexity rises. As previously stated, the researchers also point out that although random circuit sampling is a standard for computational benefit, it is not a straightforward solution to practical issues.
The advantage gap may be closing as classical supercomputing approaches for mimicking quantum circuits continue to advance. The durability of any claimed quantum advantage could be called into question by developments in tensor network algorithms and other traditional methods.
Boosting Qubit Counts and Enhancing Fidelity
According to the study, expanding the number of qubits and enhancing circuit integrity will be essential for developing real-world uses for quantum computing. The researchers point to drug discovery, machine learning, and optimisation issues as possible short-term benefits of these advancements.
Given the speed at which quantum hardware is developing, the next stage is likely to concentrate on fault tolerance and error correction, two essential components of large-scale, useful quantum computing. Global businesses and organisations, such as Google, IBM, and several Chinese research teams, are stepping up their efforts in these fields.
Global race in quantum computing
With research and development being driven by big businesses like Google, Microsoft, and IBM, investment in quantum computing has increased quickly. As Microsoft investigates topological qubits for scalable quantum systems, Google continues to concentrate on fault-tolerant quantum computing.
Read more on Majorana 1: Quantum Processor With Topological Qubits
Read more on Google Willow: From 10 Septillion Years to 5 Minutes
Alibaba has concentrated on quantum cryptography and simulation, whereas IBM has advanced superconducting quantum processors. The United States and China are still at the forefront of quantum research, with financing from the public and corporate sectors propelling developments in hardware, algorithms, and applications.
In conclusion
A Chinese quantum processor with 105 qubits, the Zuchongzhi-3, beats Google’s Sycamore, performing calculations in seconds that would take 6.4 billion years for traditional supercomputers. With high fidelity (99.90%), improved qubit connectivity, and a million times speedup over Google’s Willow, it firmly establishes China as a leader in the development of quantum computing.
