James R. Seddon

In the stabilizer circuit model of quantum computation, universality requires a resource known as magic. Here, we propose three new ways to quantify the magic of a quantum state using magic monotones, apply the monotones in the characterization of state conversions under stabilizer operations, and connect them with the classical simulation of quantum circuits. We first present a complete theory of these quantifiers for tensor products of single-qubit states, for which the monotones are all equal and all act multiplicatively, constituting the first qubit magic monotones to have this property. We use the monotones to establish several asymptotic and non-asymptotic bounds on state interconversion and distillation rates. We then relate our quantifiers directly to the runtime of classical simulation algorithms, showing that a large amount of magic is a necessary requirement for any quantum speedup. One of our classical simulation algorithms is a quasi-probability simulator with its runtime connected to a generalized notion of negativity, which is exponentially faster than all prior qubit quasi-probability simulation algorithms. We also introduce a new variant of the stabilizer rank simulation algorithm suitable for mixed states, while improving the runtime bounds for this class of simulations. Our work reveals interesting connections between quasi-probability and stabilizer rank simulators, which previously appeared to be unrelated. Generalizing the approach beyond the theory of magic states, we establish methods for the quantitative characterization of classical simulability for more general quantum resources, and use them in the resource theory of quantum coherence to connect the L1-norm of coherence with the simulation of free operations.