A survey of dynamical decoupling sequences on a programmable superconducting quantum computer

Dynamical decoupling (DD) is the judicious placement of control pulses to decouple a quantum system from its environment without the need for feedback. Error-suppression through DD sequences is well suited for noisy intermediate-scale quantum (NISQ) era quantum computers (QCs) due to its low resource overhead. In this work we update the status of previous DD surveys in light of recent advancements with cloud-based superconducting qubit devices. In particular, we use an IBM cloud QC with additional control of input pulses through the OpenPulse API. These additional controls allow us to implement certain robust and non-uniformly spaced sequences which--to our knowledge--have not been tested on cloud-based QCs. We use our experimental results to (1) find the best performing DD sequence and (2) test predictions from DD theory. For example, we address the effect of finite-pulse width errors, concatenation depth, and delay duration between pulses on DD performance.