- Sputnik International
Asia
Find top stories and features from Asia and the Pacific region. Keep updated on major political stories and analyses from Asia and the Pacific. All you want to know about China, Japan, North and South Korea, India and Pakistan, Southeast Asia and Oceania.

China’s Quantum Computer Beats Google’s Sycamore in 'Computational Supremacy', Claims New Study

CC0 / / Quantum computer
Quantum computer  - Sputnik International
Subscribe
Previously, Google boasted of having achieved quantum supremacy in a 2019 Nature paper, claiming its Sycamore device had completed a task in 200 seconds that the company claimed would take a state-of-the-art supercomputer 10,000 years to finish.

A team of researchers affiliated with several institutions in China claims to have built and tested a photonic quantum computer that has achieved quantum supremacy.

In their paper, published in the journal Science, the team led by Chinese quantum physicists Pan Jianwei and Lu Chaoyang announced that their computer, which they call Jiuzhang, after an ancient Chinese mathematical text, had performed the task while conducting Gaussian boson sampling.

​In the study, the team of the University of Science and Technology of China chose the boson sampling (GBS) - a classical simulation algorithm - to demonstrate quantum computational speed in solving well-defined tasks.

Sycamore ‘Outperformed’

Previously, only one computer has reportedly ever achieved this feat, as Sycamore, Google's quantum processor, comprising 54 qubits, was described in a 2019 Nature paper as having taken just 200 seconds to complete a task that the American multinational technology company claimed would take a state-of-the-art supercomputer 10,000 years to finish. Thus, Google claimed to have achieved quantum supremacy.

​Lu Chaoyang, a professor of the University of Science and Technology of China in Hefei, capital of East China's Anhui Province, who plays a key leading role in the research team, was quoted by the Global Times as claiming their quantum computer had multiple advantages over its Google counterpart.

Jiuzhang ostensibly outperformed Sycamore in computational speed, environmental suitability and computing power on problems with larger samples, Lu stated.

The Chinese quantum computer could find solutions to the boson-sampling problem in 200 seconds.

According to a statement cited by Global Times, Jiuzhang is 100 trillion times faster than the most powerful supercomputer of today, and 10 billion times faster than Google's Sycamore.

‘Quantum Supremacy’

Quantum computers that could potentially outperform conventional machines on specific tasks have been making the headlines of late, seeking to achieve what has come to be known as" quantum supremacy". This refers to a situation when a quantum computer can outperform conventional computers on at least one type of task.

Researchers globally are working vastly different designs to achieve this goal. Thus, Sycamore was based on qubits represented by superconducting materials.

The new research conducted by the team in China has developed a photon-based quantum computer.

Boson Sampling

The simulation algorithm of Boson sampling is a means used to calculate the output of a straight-line optical circuit with multiple inputs and outputs.

To enable this, a machine is constructed where photons are sent into a circuit in parallel. Inside, they are split by beam splitters and continue through the circuit, encountering mirrors and other beam splitters.

CC BY-SA 4.0 / Oleg Alexandrov / D-wave computer inside of the Pleiades supercomputerQuantum D-wave computer inside of the Pleiades supercomputer at NASA
China’s Quantum Computer Beats Google’s Sycamore in 'Computational Supremacy', Claims New Study - Sputnik International
Quantum D-wave computer inside of the Pleiades supercomputer at NASA

If two photons encounter the same splitter simultaneously, both unsplit photons will follow one of the paths away from the splitter in a repeated process, resulting in a distribution of numbers. The latter represent the network output. When faced with the task of calculating distributions of such a system, conventional computers baulk very quickly.

Jiuzhang, according to the study, was constructed to manage 100 inputs and 100 outputs using 300 beam splitters and 75 mirrors.

As for its future application, Jiuzhang’s super-computing is claimed to be best-suited to the needs of such areas as graph theory, machine learning and quantum chemistry, according to the team.

Newsfeed
0
To participate in the discussion
log in or register
loader
Chats
Заголовок открываемого материала