Provable classically intractable sampling with measurement-based computation in constant time

We present a constant-time measurement-based quantum computation (MQC) protocol to perform a classically intractable sampling problem. We sample from the output probability distribution of a subclass of the instantaneous quantum polynomial time circuits introduced by Bremner, Montanaro and Shepherd. In contrast with the usual circuit model, our MQC implementation includes additional randomness due to byproduct operators associated with the computation. Despite this additional randomness we show that our sampling task cannot be efficiently simulated by a classical computer. We extend previous results to verify the quantum supremacy of our sampling protocol efficiently using only single-qubit Pauli measurements.