Classical Supercomputer Challenge Disrupts Google's Quantum Supremacy Claim
A team of Chinese researchers uses a classical supercomputer to outperform Google's quantum processor on a specific benchmark. This development shows the competition between quantum and classical computing remains unresolved.
Researchers at the Chinese Academy of Sciences challenge Google's 2019 quantum supremacy claim by accomplishing the same task on an ordinary classical supercomputer. Led by Pan Zhang, the team develops a new technique that simulates the output of Google's Sycamore quantum processor in a fraction of the previously estimated time. This does not mean Google misrepresented its groundbreaking work, but it does show that the classical versus quantum computing race remains wide open.
Google's original quantum supremacy demonstration relies on finding one highly specific task that a quantum computer performs better than any classical machine. For Sycamore, that sole task is simulating a quantum computer itself. While this seems like a loophole, proving the viability of quantum methods on even a weird, narrow benchmark serves as a crucial stepping stone for the entire industry to expand its practical applications.
To achieve this classical milestone, Zhang's team models Google's 53-qubit system as a massive 3D network of tensors. They represent the 20 cycles of Sycamore's gates by extruding this grid of nodes, and then calculate the complex mathematical relationships using a cluster of 512 GPUs. This innovative tensor approach dramatically reduces the computation time, proving that classical hardware still holds surprising potential to keep pace with early quantum systems.