# A Controlled Diagnostic Study of Hardware-Induced Distortions in Hardware-Aware Training

arXiv: https://arxiv.org/abs/2605.09416 · HTML: https://mlsanity.com/paper/2605.09416/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 6 of its 19 references that are in the map.
- First posted on arXiv 10 May 2026.

## Builds on (most cited first, 6 of 6)

- [Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference](https://mlsanity.com/paper/1712.05877/) (2017)
- [Quantized Neural Networks: Training Neural Networks with Low Precision Weights and Activations](https://mlsanity.com/paper/1609.07061/) (2016)
- [Understanding Straight-Through Estimator in Training Activation Quantized Neural Nets](https://mlsanity.com/paper/1903.05662/) (2019)
- [A flexible and fast PyTorch toolkit for simulating training and inference on analog crossbar arrays](https://mlsanity.com/paper/2104.02184/) (2021)
- [Fractional Underdamped Langevin Dynamics: Retargeting SGD with Momentum under Heavy-Tailed Gradient Noise](https://mlsanity.com/paper/2002.05685/) (2020)
- [Time-Varying Convex Optimization: Time-Structured Algorithms and Applications](https://mlsanity.com/paper/2006.08500/) (2020)

## Abstract (arXiv)

Hardware-aware training (HAT) is widely used to improve the robustness of neural networks on non-ideal AI accelerators, such as analog in-memory computing (IMC) systems. However, not all hardware-induced distortions are equally compensable by training. This paper presents a diagnostic framework that models hardware non-idealities as structured perturbations of the forward operator and evaluates their compatibility with gradient-based optimization. We analyze six representative perturbation classes--read noise, variability, drift, stuck-at faults, IR-drop, and ADC discretization--and identify three key diagnostics: gradient expectation consistency, bounded gradient variance, and non-degenerate sensitivity. Our results show a clear separation between perturbations that can be compensated by HAT and those that consistently break optimization. This provides practical guidance for hardware-software co-design, clarifying which non-idealities can be addressed at the training level and which require circuit-, architecture-, or calibration-level mitigation. This study should be interpreted as a controlled empirical analysis under vanilla forward-perturbation HAT, rather than as a universal theory of hardware-aware training.