Physical realizations of quantum computers are inherently noisy. Quantum circuits run on such noisy devices are subjected to errors and inaccuracies, resulting in outcomes which strongly deviate from ideal, already for rather small quantum computations. Arguably, the most pressing challenge faced by hardware developers of quantum computers is to reduce these effects of noise and inaccuracies on the performance of quantum circuits. This is important not only for the ability to run useful quantum algorithms in the “NISQ era”, namely with the relatively small and noisy quantum computers available today and in the near future, but also for lowering the barriers and reducing the immense resources required for achieving the holy grail of quantum fault tolerance.  

Qedma’s fidelity enhancement compiler is a software product which addressed this challenge by providing error suppression and error mitigation solutions. Based on Qedma’s detailed characterization of the noisy components in the device, our compiler transforms an algorithm designed to work with ideal gates into one that is designed to work with the inaccurate gates available in the noisy quantum computer. Qedma’s compiler applies multiple runs of one or more noisy quantum circuits, and combines their outputs using classical post processing to provide the output. This results in a significant boost in the output fidelity of the given quantum circuit, for a wide variety of quantum algorithms.   

Qedma’s fidelity enhancement compiler reduces the error at the algorithmic level; it can be easily combined with Qedma’s gate optimizer, which improves the performance of the gates at the pulse level, leading to even better results.