Google Quantum AI researchers say they have achieved a major step forward in quantum error correction, reporting a 100-fold reduction in logical qubit error rates with a new surface code approach. The team says the result was demonstrated on its 105-qubit processor, marking an advance in one of the biggest technical barriers to practical quantum computing.

Quantum computers are highly sensitive to noise, which can quickly corrupt calculations. Error correction is designed to protect fragile quantum states and make longer, more reliable computations possible. According to the researchers, the new implementation improves how logical qubits are stabilized, offering a stronger path toward systems that can scale beyond experimental use.

The company’s claims, while promising, still sit within a field where independent validation and real-world scalability remain critical. Researchers across the industry have long aimed to reduce error rates enough to support useful quantum applications in chemistry, materials science, and optimization.

If confirmed and extended, the result could help move quantum computing closer to practical deployment. For now, the reported improvement adds momentum to a race in which hardware quality, error correction, and scale remain the central challenges.