Dynamically corrected entangling gates for spin qubits*

In this talk, I present a theoretical toolbox of control protocols designed explicitly to exploit the strengths of semiconductor spin qubits and circumvent their weaknesses while generating robust entangling gates.

Spin qubits have moved beyond single-qubit demonstrations of their favorable properties such as exceptionally long T₂ times and rapid gating. Now, multiple experimental groups have achieved two-qubit entangling gate operations with fidelities at or above 90%, but still short of the crucial thresholds for quantum error correcting codes. While semiconductor-based platforms are highly desirable for scalability, they universally suffer from 1/f charge noise, and generally from low-frequency magnetic noise as well, causing T₂^* to be orders of magnitude shorter than T₂. Entangling gates are particularly susceptible to this noise. I show, though, that this need not be a limitation. The separation of timescales permits efficient dynamical correction, as demonstrated by the wide use of spin echo techniques such as CPMG to extend coherence times of idle bits. The challenge is to perform dynamically corrected entangling gates, boosting gate fidelities in the current devices.

For specific, well-known experimental setups in both Si and GaAs, I show that there are combinations of simple pulse sequences and optimal operating points in the available parameter space that suffice to boost simulated two-qubit fidelities well above 99% under realistic noise modeling. These are all open-loop control techniques, but I also show a generic modular approach which allows direct integration with experimental closed loop techniques for further protection.