Measurement-free Quantum Error Correction with a Digital Entropy Pump: Part 2

As quantum computing has become viable in concept, attention has shifted toward addressing the challenges of large-scale practical computing. This includes quantum error correction at-scale, which in the current paradigm is limited in part by challenges associated with real-time measurement, decoding and control feedback with a classical processing unit. We propose a method to circumvent these challenges and eliminate any need for feedback with a classical processor for the purposes of error correction. The method transfers the entropy associated with unknown errors from data qubits to ancillas, then unconditionally resets the ancillas to their ground state in an overall process that can be recognized as a digital entropy pump. Using a stabilizer code, the entropy transfer from data qubits is achieved with gate-level instructions that extract error syndromes, decode them and apply corresponding corrections coherently all within the quantum processor. Resetting ancillas then removes this entropy from the system through the quantum hardware’s engineered dissipative coupling to its environment.

In this talk, we discuss concrete strategies for measurement-free decoding of errors on the surface code. We first introduce a simple “lookup table” method that conditionally corrects every possible syndrome. We then consider decoding via minimization of a random-bond Ising Hamiltonian with a QAOA-type algorithm. We finish by discussing future prospects and challenges for engineering an effective “Ising decoder”.