Dr. Robin Harper watched as the logical qubit’s survival rate climbed past 96%—a number that, just months earlier, seemed out of reach. On a 156-qubit IBM Quantum Heron r2 processor, Harper and his team from the University of Sydney had reengineered the delicate choreography of quantum error correction, turning a persistent stumble into a stride forward. Their breakthrough, achieved in collaboration with IBM, brings the dream of reliable, large-scale quantum computing one critical step closer to reality.
Quantum computers hold immense promise: they could unlock new materials, revolutionize drug discovery, and solve optimization problems that stump today’s supercomputers. But their fragile qubits—easily disturbed by heat, electromagnetic fields, or even time itself—have made stability a monumental challenge. Error correction is essential, yet the very act of checking for errors introduces new ones, especially during mid-circuit measurements, when qubits are briefly observed to preserve coherence. These measurements force other qubits to idle, accumulating noise and degrading performance.
The Sydney-IBM team focused on this bottleneck. Using the heavy-hex code layout of the Heron r2 processor, they designed and tested new error-correction circuits that minimized idling time during mid-circuit measurements. In memory experiments with a distance d = 3 code patch, they observed logical qubit survival rates jump from below 90% to over 96% per error-correction cycle—a dramatic improvement in quantum reliability. They also identified measurement noise as a dominant source of failure in current quantum logic operations, a finding that redirects engineering priorities toward cleaner, faster readouts.
"We wanted to identify which physical processes were limiting performance on modern quantum devices," said Dr. Harper, lead author and researcher at Sydney Nano. "What we found is that the act of measuring qubits during a calculation can itself create instability." By refining how and when those measurements occur, the team not only reduced error rates but also provided a clear benchmark for future hardware development. Their work, published in Nature Communications, offers a roadmap for scaling quantum systems with greater confidence.
This progress stems from the 2024 partnership between the University of Sydney and IBM, a collaboration aimed at tackling the core barriers to quantum scalability. With quantum error correction now better understood—and demonstrably improved—the path ahead is clearer. As engineers refine measurement speed and qubit isolation, each cycle brings us nearer to quantum machines that don’t just compute, but compute reliably.