arXiv2026
In near-term quantum computations that do not employ fault tolerant error correction, noise can proliferate rapidly, corrupting the quantum state and making results unreliable. These errors originate from both decoherence and control imprecision and the latter can manifest as coherent error that is especially detrimental. In the pre-fault tolerant setting, previous work has shown that different error suppression methods have shown promising complementary advantages but highly variable performance under different algorithmic and error model conditions. Here, we evaluate the effectiveness of several error suppression methods under varying algorithmic settings, both theoretically with numerical simulations and experimentally on a trapped-ion quantum computer. For our case study, we explore a range of output states under Grover's algorithm quantum circuits containing up to 10 qubits and 28 two-qubit gates with varying output state features. We explore the complementary effectiveness of randomized compiling and algorithm error detection, where the latter is realized via post-selection on ancillary qubits that ideally return to the ground state at the end of each circuit. In all settings, combining randomized compiling and error detection yields the largest suppression of error, indicating that these methods are most effective when combined to extend the capabilities of near-term devices for moderately deep circuits. We demonstrate for the first time significant suppression of coherent error on a trapped-ion platform, and moreover achieve this outcome via cloud access. However our results highlight that the degree of error suppression depends sensitively on the nature of the error model and the algorithm instance.