Hardware-Accelerated GNN-based Hit Filtering for the Belle II Level-1 Trigger
T. Ferber · 2025
Read Hardware-Accelerated GNN-based Hit Filtering for the Belle II Level-1 Trigger by T. Ferber (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-04731.txt.
Description
Greta Heine leads five authors putting graph neural networks on chips for the Belle II trigger. The Level-1 trigger must filter detector hits in real time at high rate. Their network reads spatial and temporal links among sense wire hits.
Quantization, pruning, and static graph building tune the model for FPGA hardware. Splitting the detector into parallel sectors lets the design cover the whole instrument. The system sustains a throughput near thirty-one point eight megahertz with latency in the hundreds of nanoseconds. It uses about a third of the chip's logic and no signal processors, leaving room to grow.
Catalog record
| CREATOR | T. Ferber |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 17 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-04731