Google recently published a paper in Nature entitled “Quantum supremacy using a programmable superconducting processor.” They claim to have used 53 qubits to solve a task that would take about 10,000 years on a classical supercomputer.
Some caution seems necessary in interpreting this claim, however. First, it does not refer to any of the practical tasks that people would like quantum computers to solve. And second, IBM claims that 2.5 days is a more realistic estimate for the time required on a classical supercomputer. It does seem that “quantum supremacy” is not quite here yet (see also what Scott Aaronson has to say).