Performance Evaluation and Optimization of the Sampling Process of an Neuropredictive SAR ADC

Spielberger A, Spitzkopf L, Schrotz AM, Pfannenmüller C, Weigel R, Franchi N (2026)

Publication Language: English

Publication Status: Accepted

Publication Type: Conference contribution, Conference Contribution

Future Publication Type: Conference contribution

Publication year: 2026

Publisher: IEEE

Event location: Busan

ISBN: 979-8-3315-8907-3

URI: https://ieeexplore.ieee.org/document/11377187

DOI: 10.1109/APCCAS67402.2025.11377187

Abstract

The functionality of the sampling process of a neuropredictive SAR ADC is analyzed and the behavior of the Neuromorphic ADC is modeled in a Python simulation and verified through measurements with different parameters. By checking the boundary values of the prediction window, clock cycles can be saved even if the prediction is incorrect. Simulative modeling enables training of the neural networks close to the hardware. The influence of the prediction window is optimized with the help of the simulation to reduce the conversion time. The maximum cycle saving for a noisy ADC is reduced from 60% to 52% for an 16b ADC with an SNR of 70 dB.

Authors with CRIS profile

How to cite

APA:

Spielberger, A., Spitzkopf, L., Schrotz, A.-M., Pfannenmüller, C., Weigel, R., & Franchi, N. (2026). Performance Evaluation and Optimization of the Sampling Process of an Neuropredictive SAR ADC. In Proceedings of the 21st IEEE Asia Pacific Conference on Circuits and Systems 2025 (APCCAS 2025). Busan, KR: IEEE.

MLA:

Spielberger, Alexander, et al. "Performance Evaluation and Optimization of the Sampling Process of an Neuropredictive SAR ADC." Proceedings of the 21st IEEE Asia Pacific Conference on Circuits and Systems 2025 (APCCAS 2025), Busan IEEE, 2026.

BibTeX: Download