Master Thesis Open Access

Training and Optimization of a Graph Neural Network for Deployment on FPGA Hardware in the Belle II Level 1 Trigger

Baptist, Frank Michael

Dublin Core Export

<?xml version='1.0' encoding='utf-8'?>
<oai_dc:dc xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd">
  <dc:creator>Baptist, Frank Michael</dc:creator>
  <dc:date>2026-02-13</dc:date>
  <dc:description>At the Belle II experiment, collisions occur at a rate higher than 200 MHz. Due to bandwidth and storage limitations, storing this amount of data for offline analysis is not feasible. That is why the experiment relies on a two-level trigger system as a preselection before passing the events to subsequent software-based processing and analysis. The first level, known as the Level 1 Trigger (L1 Trigger), is a hardware-based system that makes decisions within a few microseconds. The buffered data from the detector can only be stored for a limited time, which imposes strict latency requirements on the L1 Trigger.

One approach to enhance the L1 Trigger is the development of the GNN-ETM. It utilizes dynamic graph building based on GravNet and can predict a varying number of clusters per event using Object Condensation. The GNN-ETM is implemented on field-programmable gate array (FPGA) hardware, which allows for low-latency inference and is suitable for deployment on the L1 Trigger. The current parasitic implementation of the GNN-ETM in the L1 Trigger has shown promising results in terms of precision and latency. Nevertheless, further improvements, especially in terms of latency, are necessary to develop a potential replacement for the clustering algorithm currently deployed on the ECL L1 Trigger.

In this thesis, I present the optimization of the GNN-ETM for reduced latency, including modifications to the model architecture and training procedure.</dc:description>
  <dc:identifier>https://publish.etp.kit.edu/record/22417</dc:identifier>
  <dc:identifier>oai:publish.etp.kit.edu:22417</dc:identifier>
  <dc:language>eng</dc:language>
  <dc:relation>url:https://publish.etp.kit.edu/communities/belle2</dc:relation>
  <dc:relation>url:https://publish.etp.kit.edu/communities/etp</dc:relation>
  <dc:relation>url:https://publish.etp.kit.edu/communities/hardware</dc:relation>
  <dc:rights>info:eu-repo/semantics/openAccess</dc:rights>
  <dc:subject>GNN-ETM</dc:subject>
  <dc:subject>GNN</dc:subject>
  <dc:subject>Belle II</dc:subject>
  <dc:subject>Trigger</dc:subject>
  <dc:subject>GravNet</dc:subject>
  <dc:subject>Object Condensation</dc:subject>
  <dc:title>Training and Optimization of a Graph Neural Network for Deployment on FPGA Hardware in the Belle II Level 1 Trigger</dc:title>
  <dc:type>info:eu-repo/semantics/masterThesis</dc:type>
  <dc:type>thesis-master-thesis</dc:type>
</oai_dc:dc>
Publication date:
February 13, 2026
Keyword(s):
GNN-ETM GNN Belle II Trigger GravNet Object Condensation
Awarding University:
Karlsruhe Institute of Technology (KIT)
Report Number:
ETP-KA/2026-03
Communities:

Cite as

Export