Quantum breakthrough: new method increases logical qubit preservation to 96% on IBM Heron processor
Quantum computing is approaching the era of fault tolerance: a team of researchers, in collaboration with IBM, has achieved an impressive 96% survival rate for logical qubits on the latest 156-qubit superconducting processor, Heron r2. This is a significant step forward compared to previous results, which did not even reach 90%.
The main stumbling block for the stable operation of quantum machines remains the so-called "idle noise." In modern systems, error correction requires regular internal checks and intermediate measurements of qubits. However, during these technological pauses, the remaining components of the processor lose stability, generating new errors and negating efforts to correct previous ones.
To solve this fundamental problem, physicists completely redesigned the architecture of error correction circuits. The key innovation was a radical reduction in the time of forced computation halts. By optimizing the algorithms, scientists were able to raise the survival rate of logical qubits per single error correction cycle from less than 90% to 96% on the IBM Quantum Heron r2 processor.
It is important to emphasize that the correction process is repeated multiple times at each stage of computation. Previously, each "idle" cycle became a serious obstacle to reliable operation. Now, by minimizing these pauses, we obtain a much more stable system.
Although the result has so far been obtained under laboratory conditions on a single processor, this is a critically important achievement for the entire industry. Scalability and fault tolerance remain the main barriers to practical quantum supremacy. Recall that IBM previously announced plans to achieve the first confirmed cases of quantum advantage by the end of 2026. This breakthrough in error correction is a direct step towards achieving that ambitious goal.
Expert comment: Increasing qubit preservation to 96% is not just a number. It is a demonstration that the "idle noise" problem has an engineering solution. If this method can be scaled to larger systems, we may see an acceleration of the transition from experimental quantum computers to commercially useful machines sooner than many expect.