A proposed gold standard family of protocols for benchmarking and diagnosing elementary quantum gate operations

We propose a family of randomized benchmarking protocols as a gold standard for assessing and diagnosing error rates on elemntary one and two qubit gates. We discuss various advantages of our bespoke family of protocols relative to other standard approaches. These advantages include: reducing the number of single shot experiments and the number of distinct random sequences including; improving the accuracy and convergence of confidence intervals; accommodating arbitrary gates within SU(4); and reliably assessing the error rate and error type associated with individual gates from any universal gate set. Time-permitting, we will report results from implementing these protocols across a variety of leading platforms, including both superconducting qubits and trapped ions.