A Chinese research team has surpassed Google, building a quantum computer that completed a calculation in just over an hour that would take classical computers more than eight years to perform.
It’s the latest milestone in a line of exciting quantum computing developments across the last two years. In that time, researchers across the world have finally reached the long-sought-after ‘quantum advantage’ – the point at which quantum computing can solve a problem that would take an impractical amount of time for classical computing.
A team from Google first achieved the milestone in 2019 using superconducting qubits (which rely on the flow of current to perform computation), followed by a team from China in 2020 that upped the ante by using photonic qubits (which are based on light and have the potential for faster operation).
Now, another Chinese team (but led by the same researcher – Jian-Wei Pan at the University of Science and Technology of China in Shanghai) has outperformed Google again.
In a study published on the pre-print server ArXiv, the team demonstrated quantum advantage using superconducting qubits on a quantum processor called Zuchongzhi.
Zuchongzhi is a 2D programmable computer that can simultaneously manipulate up to 66 qubits. The new demonstration used 56 of them to tackle a computational problem designed to test the computer’s prowess – namely, sampling the output distribution of random quantum circuits. The theoretical basis for this problem is tricky to summarise, involving random matrix theory, mathematical analysis, quantum chaos, computational complexity and probability theory, but the important thing to know is that the time it takes to solve this problem scales up exponentially as more qubits are added to the system. This makes it quickly unmanageable for classical supercomputers, and therefore a suitable test bed for achieving quantum advantage.
“We estimate that the sampling task finished by Zuchongzhi in about 1.2 hours will take the most powerful supercomputer at least eight years,” the team reports in its paper.
“Our work establishes an unambiguous quantum computational advantage that is infeasible for classical computation in a reasonable amount of time. The high-precision and programmable quantum computing platform opens a new door to explore novel many-body phenomena and implement complex quantum algorithms.”
This problem was around 100 times more challenging than the one solved by Google’s Sycamore processor in 2019. While Sycamore used 54 qubits, Zuchongzhi used 56, showing that by increasing the number of qubits, a processor’s performance will improve exponentially.
These numbers fall far short of the 76 photonic qubits used in the Chinese team’s 2020 demonstration, but that processor involved a novel set-up of lasers, mirrors, prisms and photon detectors, and was not programmable like Sycamore or Zuchongzhi.