Journal Article/Contribution to a conference proceedings

FZJ-2026-05548

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Quantization Effects of Artificial Neural Networks for Embedded Edge-Computing ApplicationsElectronic Communications of the EASST

Aksoy, A.FZJ* ; Bekman, I. (Corresponding author)FZJ* ; Dimitrov, V. ; Dorosti, Q. ; Eguzo, C.FZJ* ; Fleitmann, S.FZJ* ; Hader, F.FZJ* ; Zambanini, A.FZJ* ; van Waasen, S.FZJ*

2026
Techn. Univ. Berlin

This record in other databases:  

Please use a persistent id in citations: doi:https://doi.org/10.14279/eceasst.v86.1.2799  doi:10.14279/ECEASST.V86.1.2799  doi:https://doi.org/10.14279/eceasst.v86.1.2799

Abstract: This paper examines the use of Quantized Neural Networks (QNNs) for two resource-constrained scientific applications: automated calibration of semiconductor quantum bits (qubits) and scientific particle detectors. We evaluate the trade-offs between Post-Training Quantization (PTQ), Quantization-Aware Training (QAT), and ultra-low-bit Binary Neural Networks (BNNs) with respect to latency and resource usage. Our results demonstrate that PTQ and QAT are easily implementable solutions achieving a four-fold reduction in memory usage for U-shaped CNN (U-Net) architectures, whereas BNNs are better suitable for use cases with challenging latency requirements. For the training of non-differentiable custom BNNs, we propose a novel, hardware-constrained learning approach using Genetic Algorithms (GAs). We showcase a Look-Up Table (LUT)-based BNN architecture suitable for direct conversion to Very High-Speed Integrated Circuit Hardware Description Language (VHDL) via the HCL4BNN framework. This method achieves nanosecond-scale inference latencies at 10 ns to 15 ns without requiring specialized Digital Signal Processor (DSP) or Block RAM (BRAM) resources.

Keyword(s): Signal Processing ; Triggering Systems ; Binary Neural Networks ; Neural Network Quantization ; Quantum Computing ; Edge Computing

---

Contributing Institute(s):

1. Integrated Computing Architectures (PGI-4)

Research Program(s):

1. 5215 - Towards Quantum and Neuromorphic Computing Functionalities (POF4-521) (POF4-521)
2. 5234 - Emerging NC Architectures (POF4-523) (POF4-523)

Appears in the scientific report 2026

---

Database coverage:
; ; DOAJ Seal ; SCOPUS

---