Steven T. Flammia

Gottesman-Kitaev-Preskill (GKP) codes are a promising candidate for implementing fault tolerant quantum computation in quantum harmonic oscillator systems such as superconducting resonators, optical photons and trapped ions, and in recent years theoretical and experimental evidence for their utility has steadily grown. It is known that logical Clifford operations on GKP codes can be implemented fault tolerantly using only Gaussian operations, and several theoretical investigations have illuminated their general structure. In this work, we explain how GKP Clifford gates arise as symplectic automorphisms of the corresponding GKP lattice and show how they are identified with the mapping class group of suitable genus $n$ surfaces. This correspondence introduces a topological interpretation of fault tolerance for GKP codes and motivates the connection between GKP codes (lattices), their Clifford gates, and algebraic curves, which we explore in depth. For a single-mode GKP code, we identify the space of all GKP codes with the moduli space of elliptic curves, given by the three sphere with a trefoil knot removed, and explain how logical degrees of freedom arise from the choice of a level structure on the corresponding curves. We discuss how the implementation of Clifford gates corresponds to homotopically nontrivial loops on the space of all GKP codes and show that the modular Rademacher function describes a topological invariant for certain Clifford gates implemented by such loops. Finally, we construct a universal family of GKP codes and show how it gives rise to an explicit construction of fiber bundle fault tolerance as proposed by Gottesman and Zhang for the GKP code. On our path towards understanding this correspondence, we introduce a general algebraic geometric perspective on GKP codes and their moduli spaces, which uncovers a map towards many possible routes of future research.

We demonstrate that small quantum memories, realized via quantum error correction in multi-qubit devices, can benefit substantially by choosing a quantum code that is tailored to the relevant error model of the system. For a biased noise model, with independent bit and phase flips occurring at different rates, we show that a single code greatly outperforms the well-studied Steane code across the full range of parameters of the noise model, including for unbiased noise. In fact, this tailored code performs almost optimally when compared with 10,000 randomly selected stabilizer codes of comparable experimental complexity. Tailored codes can even outperform the Steane code with realistic experimental noise, and without any increase in the experimental complexity, as we demonstrate by comparison in the observed error model in a recent 7-qubit trapped ion experiment.