Fig. 3. Parity curves for EASE and sequential gates and a 11-qubit TIQIP all-to-all connectivity diagram.

The connectivity diagram displays the ion pairs used in the experiments. The associated fidelities are computed from the amplitudes of the measured parity and populations via maximum likelihood estimation. a Parity curve for an EASE gate with five simultaneous XX interactions. We chose pulses with N_seg = 35 and gate time τ = 924.0 μs. This gate yielded an average fidelity of $88.3_{-1.6}^{+1.0}\%$ with an average deviation from the ideal fidelity of $2.3_{-1.6}^{+2.1}\%$ for the 50 non-involved pairs. b Parity curve for a series of five sequential XX interactions. We chose pulses with N_seg = 27 for each XX gate with gate time τ = 318.6 μs, which yielded an average fidelity of $92.0_{-1.4}^{+0.8}\%$ with an average deviation from the ideal fidelity of $0.9_{-1.0}^{+2.4}\%$ for the 50 non-involved pairs. To ensure a fair comparison of gate times, we made sure that the peak powers at which we executed the EASE and the sequential gates differed by no more than 0.5%. Thus, comparing the EASE and sequential-based approaches to create the same final state of all qubits, application of the EASE gate saved ~669.0 μs, i.e., 42% of the total gate time needed in the sequential approach. The quoted errors are 1σ confidence intervals from the maximum likelihood estimation. See Supplementary Note 3 for implementation details.