Mahmoud Mahajna
Design, Implementation, and Analysis of a Neuromorphic Data Converter Based on Y-Flash Analog Memory in 180nm CMOS.
Rel. Luciano Lavagno, Shahar Kvatinsky, Nicolas Wainstein. Politecnico di Torino, Corso di laurea magistrale in Ingegneria Elettronica (Electronic Engineering), 2026
Abstract
In the emerging fields of neuromorphic computing and in-memory computing, Analog-to-Digital Converters (ADCs) present a critical performance bottleneck, often consuming up to 80% of total system energy and occupying nearly 70% of the core chip area. Successive Approximation Register (SAR) ADCs are widely utilized in these domains; however, their performance is heavily bottlenecked by conventional resistive and capacitive Digital-to-Analog Converters (DACs). These traditional DACs are highly susceptible to fabrication mismatch, which severely degrades the ADC's resolution and efficiency. Mitigating these inaccuracies typically requires advanced, complex calibration techniques that further inflate area and power consumption. To address these limitations, this research proposes a Neuromorphic Data Converter (NDC) based on a loop-unrolled SAR ADC architecture.
The design features a programmable, non-volatile DAC utilizing Y-Flash memory cells
Relatori
Anno Accademico
Tipo di pubblicazione
Numero di pagine
Informazioni aggiuntive
Corso di laurea
Classe di laurea
Aziende collaboratrici
URI
![]() |
Modifica (riservato agli operatori) |
