IBM Quantum Computer Solves a Classically Intractable Problem in 15 Minutes

IBM and the University of Chicago have announced a quantum computing demonstration that meets key criteria for "quantum advantage" — evidence that a quantum computer can do something beyond the practical reach of classical computers while also proving its result is trustworthy. Using a newly designed error-correction strategy, the team encoded 70 logical qubits and completed a calculation in about 15 minutes that would take leading classical simulation methods an impractical amount of time to finish.

Quantum computers are usually tested with a benchmark called random circuit sampling (RCS), which asks a quantum machine to generate patterns too complex for a classical computer to reproduce. The trouble is that once a result becomes too complex to check with a classical computer, verifying that the quantum computer got it right becomes difficult too. The IBM and University of Chicago team built a more structured version of RCS that keeps the same level of difficulty for classical computers while letting errors be detected during the computation itself.

Logical qubits are built from several physical qubits working together to protect quantum information from errors and noise. In the demonstration, the 70 logical qubits carried out 2,415 logical two-qubit operations and 468 logical "T gates," and their error rates were 10 times lower than the underlying physical qubits' error rates. "Verification remains one of the biggest challenges in firmly establishing experimental quantum advantage," said Bill Fefferman, associate professor at the University of Chicago. His student Soumik Ghosh said the advances in verification could also unlock practical applications for future quantum computers.

"We are now firmly in the quantum advantage era," said Jay Gambetta, director of IBM Research and an IBM Fellow. The circuits and results from the experiment have been made publicly available through the Quantum Advantage Tracker.