Google Quantum AI researchers have reduced error rates in quantum state preparation by approximately 40 times using the Willow processor. The team utilized a technique called magic state cultivation to improve the precision of quantum states, which are essential for complex computations and often consume up to 95% of a computer's processing power.
The experimental method functions as a filter, where the system generates initial states, checks them for errors, and discards failed attempts. While the process achieved 99.99% accuracy for logical T-states, only 8% of attempts passed the verification. The researchers also successfully demonstrated the transfer of these states between error correction codes.
In a separate study, researchers from Google, DeepMind, and MIT proposed a modular architecture using hyperbolic error correction codes. Simulations suggest this approach could reduce the number of physical qubits required by 10 times compared to standard surface codes. The model incorporates the AlphaQubit neural network decoder to manage error detection.
Although these advancements address the high physical qubit requirements for fault-tolerant computing, they do not enable the immediate cracking of cryptographic systems. Current quantum hardware remains far from the scale needed to threaten Bitcoin or Ethereum, as experts estimate that thousands of physical qubits are required for such tasks.
