In recent months, the community has made decisive progress toward fault tolerance: with the Willow chip, Google demonstrated surface code logical memories that, as code distance increases, reduce error exponentially (“below threshold”). Concretely: a distance-7 logical built by 101 physical qubits achieved 0,143% error per cycle and a lifetime beyond break-even, with real-time decoding at ~63 μsThis is tangible evidence that scaling up can yield increasingly reliable logical qubits—an essential prerequisite for useful algorithms. Nature
In parallel, Microsoft and Quantinuum demonstrated logical qubits with record reliability (14,000 error-free experiments on H-Series), supporting the transition to the so-called “Level 2, Resilient” stage and pointing to thresholds (100–1000 reliable logical qubits) as the point where one can begin to speak of scientific—and eventually commercial—advantage. On the academic–industrial front, other works have demonstrated four-logical-qubit entanglement beyond break-even, signaling that logical computation is moving out of the fundamental physics lab into an engineering regime. The Official Microsoft BlogPhysical Review
Why it matters: without logical qubits that improve as scale increases, no architecture (superconductors, trapped ions, neutral atoms, or photonics) will ever run deep algorithms. These results set a new benchmark of maturity for error correction and for the hardware–software co-design needed in the post-NISQ era. NatureThe Official Microsoft Blog
Leave a Reply