IBM and The University of Chicago Demonstrate Quantum Advantage, Establishing Trusted Quantum Computation on Logical Circuits
| Source: IBM Newsroom AI
Tags: IBM, University of Chicago, quantum advantage, logical qubits, quantum error correction, random circuit sampling
IBM and the University of Chicago demonstrated quantum advantage on logical circuits using 70 error-corrected qubits — running 2,415 two-qubit operations and 468 T gates in ~15 minutes on a task where classical methods are infeasible, with a novel circuit structure that simultaneously proves hardness and enables fidelity verification.
Details
IBM and University of Chicago researchers published 'Sampling hard circuits with verifiably high fidelity,' demonstrating quantum advantage through a new approach to random circuit sampling (RCS). The key innovation: a structured circuit construction that retains the computational hardness of RCS while also enabling error detection during the computation — solving the longstanding tension between problem difficulty and result verifiability.\n\nThe experiment encoded 70 logical qubits — one of the largest logical quantum computing demonstrations to date. The circuits ran 2,415 logical two-qubit operations and 468 logical T gates (T gates measure circuit complexity and are especially difficult for classical simulation). The computation completed in approximately 15 minutes; an equivalent classical computation would be infeasible with current methods.\n\nStandard RCS has been used to benchmark quantum supremacy since Google's 2019 result, but the fundamental challenge has always been verification: as circuits scale, it becomes impossible to check the quantum result against a classical simulation. The UChicago-IBM construction breaks this barrier by embedding verifiable structure that can detect errors without giving the classical simulator a foothold.\n\nBill Fefferman (Associate Professor, UChicago) and PhD student Soumik Ghosh led the academic side. Results are openly released on IBM's Quantum Advantage Tracker, enabling external challenge.