# Biased Local SGD for Efficient Deep Learning on Heterogeneous Systems

arXiv: https://arxiv.org/abs/2508.08540 · HTML: https://mlsanity.com/paper/2508.08540/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 16 of its 31 references that are in the map.
- First posted on arXiv 12 Aug 2025.

## Builds on (most cited first, 16 of 16)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://mlsanity.com/paper/1910.01108/) (2019)
- [Wide Residual Networks](https://mlsanity.com/paper/1605.07146/) (2016)
- [Character-level Convolutional Networks for Text Classification](https://mlsanity.com/paper/1509.01626/) (2015)
- [Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization](https://mlsanity.com/paper/2007.07481/) (2020)
- [Local SGD Converges Fast and Communicates Little](https://mlsanity.com/paper/1805.09767/) (2018)
- [Not All Samples Are Created Equal: Deep Learning with Importance Sampling](https://mlsanity.com/paper/1803.00942/) (2018)
- [Tighter Theory for Local SGD on Identical and Heterogeneous Data](https://mlsanity.com/paper/1909.04746/) (2019)
- [Active Bias: Training More Accurate Neural Networks by Emphasizing High Variance Samples](https://mlsanity.com/paper/1704.07433/) (2017)
- [DistDGL: Distributed Graph Neural Network Training for Billion-Scale Graphs](https://mlsanity.com/paper/2010.05337/) (2020)
- [Local SGD with Periodic Averaging: Tighter Analysis and Adaptive Synchronization](https://mlsanity.com/paper/1910.13598/) (2019)
- [Addressing "Documentation Debt" in Machine Learning Research: A Retrospective Datasheet for BookCorpus](https://mlsanity.com/paper/2105.05241/) (2021)
- [STL-SGD: Speeding Up Local SGD with Stagewise Communication Period](https://mlsanity.com/paper/2006.06377/) (2020)
- [HetSeq: Distributed GPU Training on Heterogeneous Infrastructure](https://mlsanity.com/paper/2009.14783/) (2020)
- [FedADP: Unified Model Aggregation for Federated Learning with Heterogeneous Model Architectures](https://mlsanity.com/paper/2505.06497/) (2025)

## Abstract (arXiv)

Most parallel neural network training methods assume homogeneous computing resources. For example, synchronous data-parallel SGD suffers from significant synchronization overhead under heterogeneous workloads, often forcing practitioners to rely only on the fastest devices (e.g., GPUs). In this work, we study local SGD for efficient parallel training on heterogeneous systems. We show that intentionally introducing bias in data sampling and model aggregation can effectively harmonize slower CPUs with faster GPUs. Our extensive empirical results demonstrate that a carefully controlled bias significantly accelerates local SGD while achieving comparable or even higher accuracy than synchronous SGD under the same epoch budget. For instance, our method trains ResNet20 on CIFAR-10 with 2 CPUs and 8 GPUs up to 32x faster than synchronous SGD, with nearly identical accuracy. These results provide practical insights into how to flexibly utilize diverse compute resources for deep learning.