GreenPeas: Unlocking adaptive quantum error correction with just-in-time decoding hypergraphs

Abstract

Circuit-level decoders are essential for the realisation of low-overhead fault-tolerant quantum computing. However, they rely on complex hypergraphs that are traditionally compiled ahead-of-time. This static approach introduces a significant bottleneck for an emerging class of adaptive circuits, where the structure is modified during execution based on mid-circuit measurement outcomes. Pre-compiling hypergraphs for all possible circuit branches would incur an exponential memory cost, rendering current tools impractical for these workloads. Hence, we introduce GreenPeas, a just-in-time compiler for decoding hypergraphs. By lowering the realised circuit to a space-time error propagation graph, GreenPeas decomposes Stim's backtracking algorithm for error analysis into two sequentially dependent, internally parallelisable stages: (1) mapping physical errors to their corresponding equivalence classes, and (2) aggregating error probabilities within each class. Evaluated on surface and bivariate bicycle code memory circuits without user-annotated repeat blocks, GreenPeas achieves a geometric mean speedup of 13.2x over Stim using a high-end GPU. This speedup carries over to the adaptive regime, unlocking circuit-level decoding of [[4,2,2]]-concatenated surface code memories with adaptive syndrome measurements -- a capability previously restricted to less accurate phenomenological decoders -- yielding 6.7x lower logical error rate and 4.5x lower decoding latency at a representative outer code distance of 10.

Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv.