DAQ-BP: Degeneracy-Aware Quantized Belief Propagation for Real-Time Decoding of Quantum LDPC Codes

Abstract Quantum low-density parity-check (qLDPC) codes, particularly bivariate bicycle codes, offer substantial reductions in physical qubit overhead compared to surface codes. However, real-time syndrome extraction in superconducting architectures demands sub-microsecond decoding latency to avoid backlog under continuous syndrome extraction cycles (~1 µs). Standard belief propagation with ordered statistics decoding (BP-OSD) achieves low logical error rates but relies on O(N³) post-processing matrix inversion, creating a severe latency and computational bottleneck. In this work, we present Degeneracy-Aware Quantized Belief Propagation (DAQ-BP), an integer-quantized decoding architecture designed for hardware implementation. DAQ-BP evaluates 5-bit integer message polarities without DSP multipliers, eliminating over 93% of O(N³) matrix inversions when deployed as a hardware front-end to BP-OSD on the IBM [[144, 12, 12]] Gross code at physical error rates p ≤ 0.01, maintaining identical logical error suppression (0.00% LER) with an estimated mean latency of 10.5 ns (1.05 iterations at 400 MHz) and bounded tail latency of 350 ns (35 iterations). Key Contributions 5-Bit Hardware Quantization: Replaces transcendental continuous hyperbolic check updates with static 32-entry lookup tables and bit-shifts, requiring zero DSP multipliers. Dynamic Damping & Momentum Kick: Dynamic shift-based damping and anti-trapping sign inversion resolve symmetric degeneracy and avoid cyclic trapping sets on bivariate bicycle graphs. O(N³) Inversion Elimination: Resolves 93.1% to 98.8% of non-zero syndromes before OSD invocation at operational error rates. Synthesizable RTL & Open Reproduction: Complete SystemVerilog node processing units, Python simulation pipelines, and interactive Jupyter validation notebooks. Code Repository Deeven-Seru/daq-bp — GitHub

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-18
DOI
https://doi.org/10.5281/zenodo.24333037
Primary Topic
Quantum Computing Algorithms and Architecture
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

DAQ-BP: Degeneracy-Aware Quantized Belief Propagation for Real-Time Decoding of Quantum LDPC Codes

Deeven Seru
1 links
Zenodo (CERN European Organization for Nuclear Research)
Quantum Computing Algorithms and Architecture
preprint

DAQ-BP: Degeneracy-Aware Quantized Belief Propagation for Real-Time Decoding of Quantum LDPC Codes

Deeven Seru
preprint en

Abstract

Abstract Quantum low-density parity-check (qLDPC) codes, particularly bivariate bicycle codes, offer substantial reductions in physical qubit overhead compared to surface codes. However, real-time syndrome extraction in superconducting architectures demands sub-microsecond decoding latency to avoid backlog under continuous syndrome extraction cycles (~1 µs). Standard belief propagation with ordered statistics decoding (BP-OSD) achieves low logical error rates but relies on O(N³) post-processing matrix inversion, creating a severe latency and computational bottleneck. In this work, we present Degeneracy-Aware Quantized Belief Propagation (DAQ-BP), an integer-quantized decoding architecture designed for hardware implementation. DAQ-BP evaluates 5-bit integer message polarities without DSP multipliers, eliminating over 93% of O(N³) matrix inversions when deployed as a hardware front-end to BP-OSD on the IBM [[144, 12, 12]] Gross code at physical error rates p ≤ 0.01, maintaining identical logical error suppression (0.00% LER) with an estimated mean latency of 10.5 ns (1.05 iterations at 400 MHz) and bounded tail latency of 350 ns (35 iterations). Key Contributions 5-Bit Hardware Quantization: Replaces transcendental continuous hyperbolic check updates with static 32-entry lookup tables and bit-shifts, requiring zero DSP multipliers. Dynamic Damping & Momentum Kick: Dynamic shift-based damping and anti-trapping sign inversion resolve symmetric degeneracy and avoid cyclic trapping sets on bivariate bicycle graphs. O(N³) Inversion Elimination: Resolves 93.1% to 98.8% of non-zero syndromes before OSD invocation at operational error rates. Synthesizable RTL & Open Reproduction: Complete SystemVerilog node processing units, Python simulation pipelines, and interactive Jupyter validation notebooks. Code Repository Deeven-Seru/daq-bp — GitHub

Zenodo (CERN European Organization for Nuclear Research)
Quantum Computing Algorithms and Architecture
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon