Tim Chan (Jun 05 2024).
Abstract: We design Snowflake, a quantum error correction decoder that runs in a streaming fashion and is capable of a simple, local implementation. In doing so we propose a new method for general stream decoding that eliminates the processing overhead due to window overlap in existing windowing methods. As a first study, we test our local implementation of Snowflake on the surface code under circuit-level noise. It recovers roughly 2/3 the accuracy threshold of the Union-Find decoder adapted with a windowing method, with a better mean runtime scaling: subquadratic as opposed to cubic in code distance $d$. We discuss how Snowflake may be implemented on a 2D chip and decode not just quantum memory but lattice surgery-based computation.