mlsanityData to 7 Oct 2026

Hardware-aware training for large-scale and diverse deep learning inference workloads using in-memory computing-based accelerators

arXiv 2302.08469 ↗First posted 16 Feb 2023cs.LG cs.ETcited by 42

Not on the map. Why?

Abstract (arXiv)

Analog in-memory computing (AIMC) -- a promising approach for energy-efficient acceleration of deep learning workloads -- computes matrix-vector multiplications (MVMs) but only approximately, due to nonidealities that often are non-deterministic or nonlinear. This can adversely impact the achievable deep neural network (DNN) inference accuracy as compared to a conventional floating point (FP) implementation. While retraining has previously been suggested to improve robustness, prior work has explored only a few DNN topologies, using disparate and overly simplified AIMC hardware models. Here, we use hardware-aware (HWA) training to systematically examine the accuracy of AIMC for multiple common artificial intelligence (AI) workloads across multiple DNN topologies, and investigate sensitivity and robustness to a broad set of nonidealities. By introducing a new and highly realistic AIMC crossbar-model, we improve significantly on earlier retraining approaches. We show that many large-scale DNNs of various topologies, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), and transformers, can in fact be successfully retrained to show iso-accuracy on AIMC. Our results further suggest that AIMC nonidealities that add noise to the inputs or outputs, not the weights, have the largest impact on DNN accuracy, and that RNNs are particularly robust to all nonidealities.

Key facts. Cited by 42 papers in the map (data through 7 Oct 2026). Builds on 32 of its 111 references that are in the map. First posted on arXiv 16 Feb 2023.

Builds on

32 of 111 references.

References · 32

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  4. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  5. 2016Densely Connected Convolutional Networkscited by 6,691
  6. 2018GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understandingcited by 4,854
  7. 2016Wide Residual Networkscited by 3,057
  8. 2015Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Codingcited by 2,291
  9. 2019SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognitioncited by 1,355
  10. 2016XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networkscited by 829
  11. 2018Quantizing deep convolutional networks for efficient inference: A whitepapercited by 507
  12. 2018PACT: Parameterized Clipping Activation for Quantized Neural Networkscited by 456
20 more
  1. 2021A White Paper on Neural Network Quantizationcited by 453
  2. 2017Efficient Processing of Deep Neural Networks: A Tutorial and Surveycited by 427
  3. 2016Quantized Neural Networks: Training Neural Networks with Low Precision Weights and Activationscited by 413
  4. 2013Maxout Networkscited by 186
  5. 2022Compute Trends Across Three Eras of Machine Learningcited by 161
  6. 2021Deep physical neural networks enabled by a backpropagation algorithm for arbitrary physical systemscited by 125
  7. 2018Parametric Noise Injection: Trainable Randomness to Improve Deep Neural Network Robustness against Adversarial Attackcited by 123
  8. 2013Dropout Training as Adaptive Regularizationcited by 89
  9. 2017Regularizing Deep Neural Networks by Noise: Its Interpretation and Optimizationcited by 56
  10. 2021A flexible and fast PyTorch toolkit for simulating training and inference on analog crossbar arrayscited by 38
  11. 2016Acceleration of Deep Neural Network Training with Resistive Cross-Point Devicescited by 28
  12. 2017Training Deep Convolutional Neural Networks with Resistive Cross-Point Devicescited by 17
  13. 2018RxNN: A Framework for Evaluating Deep Neural Networks on Resistive Crossbarscited by 17
  14. 2020Noisy Machines: Understanding Noisy Neural Networks and Enhancing Robustness to Analog Hardware Errors Using Distillationcited by 17
  15. 2021AI Accelerator Survey and Trendscited by 17
  16. 2016Whiteout: Gaussian Adaptive Noise Regularization in Deep Neural Networkscited by 8
  17. 2017Embedding-Based Speaker Adaptive Training of Deep Neural Networkscited by 7
  18. 2020TxSim:Modeling Training of Deep Neural Networks on Resistive Crossbar Systemscited by 7
  19. 2019Training large-scale ANNs on simulated resistive crossbar arrayscited by 5
  20. 2018Efficient ConvNets for Analog Arrayscited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 8; 2024: 14; 2025: 10; 2026: 10.

42 papers.

Sort:

By area:

  1. 2024Training of Physical Neural Networkscited by 47
  2. 2024Memory Is All You Need: An Overview of Compute-in-Memory Architectures for Accelerating Large Language Model Inferencecited by 14
  3. 2023Fast offset corrected in-memory trainingcited by 11
  4. 2023AnalogNAS: A Neural Network Design Framework for Accurate Inference with Analog In-Memory Computingcited by 10
  5. 2023Using the IBM Analog In-Memory Hardware Acceleration Kit for Neural Network Training and Inferencecited by 10
  6. 2024Towards Exact Gradient-based Training on Analog In-memory Computingcited by 6
  7. 2023Examining the Role and Limits of Batchnorm Optimization to Mitigate Diverse Hardware-noise in In-memory Computingcited by 5
  8. 2023Mixture of Tokens: Continuous MoE through Cross-Example Aggregationcited by 5
  9. 2024ClipFormer: Key-Value Clipping of Transformers on Memristive Crossbars for Write Noise Mitigationcited by 4
  10. 2025Analog In-memory Training on General Non-ideal Resistive Elements: The Impact of Response Functionscited by 4
  11. 2023Measurement-driven neural-network training for integrated magnetic tunnel junction arrayscited by 3
  12. 2024DIMAT: Decentralized Iterative Merging-And-Training for Deep Learning Modelscited by 3
30 more
  1. 2024Towards training digitally-tied analog blocks via hybrid gradient computationcited by 3
  2. 2025Analog Foundation Modelscited by 3
  3. 2024TReX- Reusing Vision Transformer's Attention for Efficient Xbar-based Computingcited by 2
  4. 2024When In-memory Computing Meets Spiking Neural Networks -- A Perspective on Device-Circuit-System-and-Algorithm Co-designcited by 1
  5. 2024Efficient transformer adaptation for analog in-memory computing via low-rank adapterscited by 1
  6. 2024Harnessing Nonidealities in Analog In-Memory Computing Circuits: A Physical Modeling Approach for Neuromorphic Systemscited by 1
  7. 2025Onboard Optimization and Learning: A Surveycited by 1
  8. 2025Rapid yet accurate Tile-circuit and device modeling for Analog In-Memory Computingcited by 1
  9. 2025AnalogNAS-Bench: A NAS Benchmark for Analog In-Memory Computingcited by 1
  10. 2025Memristor-Based Neural Network Accelerators for Space Applications: Enhancing Performance with Temporal Averaging and SIRENscited by 1
  11. 2023Hyperspectral In-Memory Computing with Optical Frequency Combs and Programmable Optical Memoriescited by 0
  12. 2023U-SWIM: Universal Selective Write-Verify for Computing-in-Memory Neural Acceleratorscited by 0
  13. 2024Learning-to-learn enables rapid learning with phase-change memory-based in-memory computingcited by 0
  14. 2024Training a multilayer dynamical spintronic network with standard machine learning tools to perform time series classificationcited by 0
  15. 2024Retro-li: Small-Scale Retrieval Augmented Generation Supporting Noisy Similarity Searches and Domain Shift Generalizationcited by 0
  16. 2024Kernel Approximation using Analog In-Memory Computingcited by 0
  17. 2025Scaling of hardware-compatible perturbative training algorithmscited by 0
  18. 2025Guidance and Control Neural Network Acceleration using Memristorscited by 0
  19. 2025MDM: Manhattan Distance Mapping of DNN Weights for Parasitic-Resistance-Resilient Memristive Crossbarscited by 0
  20. 2025Personalized Spiking Neural Networks with Ferroelectric Synapses for EEG Signal Processingcited by 0
  21. 2026Incorruptible Neural Networks: Training Models that can Generalize to Large Internal Perturbationscited by 0
  22. 2026Robust Heterogeneous Analog-Digital Computing for Mixture-of-Experts Models with Theoretical Generalization Guaranteescited by 0
  23. 2026Joint Hardware-Workload Co-Optimization for In-Memory Computing Acceleratorscited by 0
  24. 2026Quantum Tunneling-Aware Machine Learning: Physics-Derived Noise Models for Robust Deploymentcited by 0
  25. 2026ReRAM-aware Model Finetuning addressing I-V Non-linearity and Retention Errorscited by 0
  26. 2026Selective KV Cache Protection for Noise-Resilient LLM Inference on Analog Compute-In-Memory Systemscited by 0
  27. 2026Large Models for Small Devices: Recent Advances and Empirical Analysis of Edge AI Deploymentcited by 0
  28. 2026RACE-AIMC: Selective Inference for Heterogeneous Analog In-Memory Accelerators at the Edgecited by 0
  29. 2026IMC-CLINIC: Coupled Loss-Informed Newton Iterations for Clipping in Analog In-Memory Computingcited by 0
  30. 2026Making Analog Training Scale: Co-Designing Mapping, Optimizer, and Converterscited by 0