Modeling Memristor-Based Neural Networks with Manhattan Update: Trade-offs in Learning Performance and Energy Consumption
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866917064295317504 |
|---|---|
| author | Quiñonez, Walter Sánchez, María José Rubi, Diego |
| author_facet | Quiñonez, Walter Sánchez, María José Rubi, Diego |
| contents | We present a systematic study of memristor based neural networks trained with the hardware-friendly Manhattan update rule, focusing on the trade offs between learning performance and energy consumption. Using realistic models of potentiation/depression (P/D) curves, we evaluate the impact of nonlinearity (NLI), conductance range, and number of accessible levels on both a single perceptron (SP) and a deep neural network (DNN) trained on the MNIST dataset. Our results show that SPs tolerate P/D nonlinearity up to NLI $\leq 0.01$, while DNNs require stricter conditions of NLI $\leq$ 0.001 to preserve accuracy. Increasing the number of discrete conductance states improves convergence, effectively acting as a finer learning rate. We further propose a strategy where one memristor of each differential pair is fixed, reducing redundant memristor conductance updates. This approach lowers training energy by nearly 50% in DNN with little to no loss in accuracy. Our findings highlight the importance of device algorithm codesign in enabling scalable, low power neuromorphic hardware for edge AI applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_03858 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Modeling Memristor-Based Neural Networks with Manhattan Update: Trade-offs in Learning Performance and Energy Consumption Quiñonez, Walter Sánchez, María José Rubi, Diego Mesoscale and Nanoscale Physics We present a systematic study of memristor based neural networks trained with the hardware-friendly Manhattan update rule, focusing on the trade offs between learning performance and energy consumption. Using realistic models of potentiation/depression (P/D) curves, we evaluate the impact of nonlinearity (NLI), conductance range, and number of accessible levels on both a single perceptron (SP) and a deep neural network (DNN) trained on the MNIST dataset. Our results show that SPs tolerate P/D nonlinearity up to NLI $\leq 0.01$, while DNNs require stricter conditions of NLI $\leq$ 0.001 to preserve accuracy. Increasing the number of discrete conductance states improves convergence, effectively acting as a finer learning rate. We further propose a strategy where one memristor of each differential pair is fixed, reducing redundant memristor conductance updates. This approach lowers training energy by nearly 50% in DNN with little to no loss in accuracy. Our findings highlight the importance of device algorithm codesign in enabling scalable, low power neuromorphic hardware for edge AI applications. |
| title | Modeling Memristor-Based Neural Networks with Manhattan Update: Trade-offs in Learning Performance and Energy Consumption |
| topic | Mesoscale and Nanoscale Physics |
| url | https://arxiv.org/abs/2511.03858 |