Recent advances in quantum computing have revealed that Google’s 67-qubit Sycamore processor can outperform the fastest classical supercomputers. This advance, detailed in a study published in the journal Nature on October 9, 2024, points to a new phase in quantum computing known as the “weak noise phase”.

Understanding the low noise phase

The research, led by Alexis Morvan of Google Quantum AI, shows how quantum processors can enter this stable computationally complex phase. During this phase, the Sycamore chip is capable of performing calculations that exceed the performance capabilities of traditional supercomputers. According to a Google representative, this discovery represents a significant step toward real-world applications for quantum technology that cannot be replicated by classical computers.

The role of Qubits in quantum computing

Quantum computers use qubits, which use the principles of quantum mechanics to perform calculations in parallel. This is in stark contrast to classical computing, where bits process information sequentially. The exponential power of qubits allows quantum machines to solve problems in seconds that would take classical computers thousands of years. However, qubits are very sensitive to interference, leading to a higher failure rate; for example, about 1 in 100 qubits can fail, compared to the incredibly low failure rate of 1 in a billion billion bits in classical systems.

Overcoming the Challenge: Noise and Error Correction

Despite its potential, quantum computing faces significant challenges, primarily noise that affects the performance of qubits. To achieve “quantum supremacy,” efficient error correction methods are needed, especially as the number of qubits increases, according to a report in LiveScience. Currently, the largest quantum machines have around 1,000 qubits, and scaling up presents complex technical hurdles.

Experiment: Random sampling circle

In a recent experiment, Google researchers used a technique called random circuit sampling (RCS) to evaluate the performance of a two-dimensional network of superconducting qubits. RCS serves as a benchmark for comparing the capabilities of quantum computers with classical supercomputers and is considered one of the most challenging benchmarks in quantum computing.

The findings showed that by manipulating noise levels and controlling quantum correlations, the researchers could shift qubits into a “low-noise phase.” In this state, the calculations became sufficiently complex, showing that the Sycamore chip can outperform classical systems.