Mitigating the Impact of Retention Loss on Inference Accuracy in 65 nm Single-Poly Floating-Gate Analog In-Memory Computing

2026/07/27 by Mirko Brazzini, Giulio Filippeschi, Alessandro Catania +2
#cs.AR #cs.NE #physics.app-ph

paper · pdf

Abstract

We show with experiments and system-level simulations that it is possible to successfully mitigate the impact of retention loss on inference accuracy degradation by using both circuit-level compensation techniques and batch normalization recalibration at the algorithmic level. Experiments are performed on a single-poly floating-gate (FG) analog non-volatile memory array for analog in-memory computing fabricated in a standard 65 nm CMOS. We use a model of retention-loss statistics calibrated with experiments to evaluate the system-level impact on neural network models such as VGG-10/CIFAR-10 and WideResNet-28-10/CIFAR-100. We show that, after 60 days since programming, combined mitigation techniques enable to recover the baseline inference accuracy within 2-4%

Citations

• Temperature-Resilient Analog Neuromorphic Chip in Single-Polysilicon CMOS Technology
• Roadmap to Neuromorphic Computing with Emerging Technologies
• Using Floating Gate Memory to Train Ideal Accuracy Neural Networks
• Improving 3D NAND Flash Memory Lifetime by Tolerating Early Retention Loss and Process Variation
• Understanding Batch Normalization
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

Related

• Accuracy and Resiliency of Analog Compute-in-Memory Inference Engines
• CorrectNet: Robustness Enhancement of Analog In-Memory Computing for Neural Networks by Error Suppression and Compensation
• Analog Foundation Models
• Rapid yet accurate Tile-circuit and device modeling for Analog In-Memory Computing
• Fundamental Limits on Energy-Delay-Accuracy of In-memory Architectures in Inference Applications