Researchers at Google have unveiled a groundbreaking quantum computer called Willow, which can perform random circuit sampling (RCS) benchmarks far faster than existing supercomputers.
This development marks a crucial advancement in how quantum computers manage noise, a major factor impacting their performance. Google aims for Willow’s technology to eventually outpace supercomputers in practical, real-world scenarios—not just in benchmarks. In a tweet, CEO Sundar Pichai referred to Willow as a state-of-the-art quantum computing chip, claiming it can drastically reduce errors. He shared that Willow completed a specific computation in less than five minutes, while a leading supercomputer would take over 1,025 years to do the same—well beyond the universe’s age.
Pichai highlighted Willow’s potential contributions to fields like drug discovery, fusion energy, and battery design. Hartmut Neven, founder and leader of Google Quantum, discussed their findings published in Nature. He noted that as Willow utilizes more qubits, its error rate decreases.
Neven explained that they tested increasingly larger grids of qubits—from 3×3 to 5×5 and up to 7×7. With each increase, they applied the latest advancements in quantum error correction, essentially halving the error rate each time. He stressed that this exponential reduction is vital for improving the computer’s capabilities.
Willow stands out as a key example of real-time error correction in superconducting quantum systems, a critical milestone for practical applications. Neven pointed out that if errors are not corrected quickly, they can compromise computations before they finish.
He described the team’s achievement as a “beyond breakeven” demonstration, indicating that the lifetime of their qubit arrays exceeds that of the individual qubits, a clear sign that their error correction efforts are yielding tangible improvements. Neven called Willow the most persuasive prototype for scalable logical qubits so far and emphasized its potential for executing practical algorithms that conventional computers cannot handle.
While he acknowledged that the benchmark used isn’t directly applicable to real-world situations, the research team’s focus remains on developing algorithms that break the barriers of classical computing and solve commercially relevant problems. Neven mentioned that their simulations of quantum systems have led to new discoveries but are still within the capabilities of classical computers.