IBM and University of Chicago researchers have demonstrated a quantum computation that exceeds the practical capabilities of leading classical simulation methods, completing the task in roughly 15 minutes. The experiment utilized 70 error-corrected logical qubits and included a method to verify the accuracy of the quantum computer's output. This achievement, detailed in a paper titled "Sampling hard circuits with verifiably high fidelity," represents one of the largest demonstrations of logical quantum computing reported to date.
The research addresses a longstanding challenge in quantum computing: how to establish confidence in results when the problem becomes too complex for classical computers to reproduce. Previous attempts to demonstrate quantum advantage often relied on benchmarks like random circuit sampling (RCS), which involves generating complex patterns that classical computers cannot efficiently replicate. However, verifying the correctness of the quantum computer's answer became increasingly difficult as the computational complexity grew.
To overcome this, the IBM and University of Chicago team developed a structured alternative to RCS. This new approach maintains the computational difficulty associated with random circuit sampling while introducing a structure that allows researchers to detect errors during the computation. According to Bill Fefferman, an Associate Professor at the University of Chicago and a co-author of the paper, this method increases confidence that the quantum computer is solving a genuinely difficult computational problem.
The experiment involved executing 2,415 logical two-qubit operations and 468 logical "T gates," which are metrics quantifying the complexity of a quantum circuit. The use of error correction techniques was central to the success of the computation. Logical qubits are encoded to protect quantum information from errors and noise, unlike individual physical qubits. The encoded design significantly improved reliability, with effective logical error rates reported to be 10 times lower than the underlying physical error rates. This allowed the circuit to maintain high fidelity even while performing a large number of quantum operations.
Jay Gambetta, Director of IBM Research and an IBM Fellow, stated that the company is now "firmly in the quantum advantage era." He noted that this experiment demonstrates a quantum computation beyond the practical reach of classical computers, with a statistically confident lower bound on its execution fidelity. The researchers found that many leading classical simulation methods would face prohibitive runtimes when attempting the same task, which the IBM quantum computer completed in approximately 15 minutes.
The circuits and experimental results from this research have been made publicly available through the Quantum Advantage Tracker. This step aims to provide transparency and allow other researchers to examine and build upon the findings. While this achievement marks a significant step, experts note that the era of general-purpose fault-tolerant quantum computing is still some years away. IBM's roadmap targets the delivery of its Starling system, a large-scale, fault-tolerant quantum computer with 200 logical qubits, by 2029.