# FreshGNN: Reducing Memory Access via Stable Historical Embeddings for Graph Neural Network Training

arXiv: https://arxiv.org/abs/2301.07482 · HTML: https://mlsanity.com/paper/2301.07482/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 35 of its 57 references that are in the map.
- First posted on arXiv 18 Jan 2023.

## Builds on (most cited first, 30 of 35)

- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Graph Attention Networks](https://mlsanity.com/paper/1710.10903/) (2017)
- [Inductive Representation Learning on Large Graphs](https://mlsanity.com/paper/1706.02216/) (2017)
- [Neural Message Passing for Quantum Chemistry](https://mlsanity.com/paper/1704.01212/) (2017)
- [Fast Graph Representation Learning with PyTorch Geometric](https://mlsanity.com/paper/1903.02428/) (2019)
- [Open Graph Benchmark: Datasets for Machine Learning on Graphs](https://mlsanity.com/paper/2005.00687/) (2020)
- [Modeling Relational Data with Graph Convolutional Networks](https://mlsanity.com/paper/1703.06103/) (2017)
- [Graph Convolutional Neural Networks for Web-Scale Recommender Systems](https://mlsanity.com/paper/1806.01973/) (2018)
- [GraphSAINT: Graph Sampling Based Inductive Learning Method](https://mlsanity.com/paper/1907.04931/) (2019)
- [FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling](https://mlsanity.com/paper/1801.10247/) (2018)
- [Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks](https://mlsanity.com/paper/1905.07953/) (2019)
- [Graph Neural Networks in Recommender Systems: A Survey](https://mlsanity.com/paper/2011.02260/) (2020)
- [OGB-LSC: A Large-Scale Challenge for Machine Learning on Graphs](https://mlsanity.com/paper/2103.09430/) (2021)
- [Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters](https://mlsanity.com/paper/2008.08692/) (2020)
- [Design Space for Graph Neural Networks](https://mlsanity.com/paper/2011.08843/) (2020)
- [Layer-Dependent Importance Sampling for Training Deep and Large Graph Convolutional Networks](https://mlsanity.com/paper/1911.07323/) (2019)
- [DistDGL: Distributed Graph Neural Network Training for Billion-Scale Graphs](https://mlsanity.com/paper/2010.05337/) (2020)
- [Decoupling the Depth and Scope of Graph Neural Networks](https://mlsanity.com/paper/2201.07858/) (2022)
- [GNNAutoScale: Scalable and Expressive Graph Neural Networks via Historical Embeddings](https://mlsanity.com/paper/2106.05609/) (2021)
- [DistGNN: Scalable Distributed Training for Large-Scale Graph Neural Networks](https://mlsanity.com/paper/2104.06700/) (2021)
- [Poseidon: An Efficient Communication Architecture for Distributed Deep Learning on GPU Clusters](https://mlsanity.com/paper/1706.03292/) (2017)
- [GNNAdvisor: An Adaptive and Efficient Runtime System for GNN Acceleration on GPUs](https://mlsanity.com/paper/2006.06608/) (2020)
- [Dorylus: Affordable, Scalable, and Accurate GNN Training with Distributed CPU Servers and Serverless Threads](https://mlsanity.com/paper/2105.11118/) (2021)
- [BGL: GPU-Efficient GNN Training by Optimizing Graph Data I/O and Preprocessing](https://mlsanity.com/paper/2112.08541/) (2021)
- [PipeGCN: Efficient Full-Graph Training of Graph Convolutional Networks with Pipelined Feature Communication](https://mlsanity.com/paper/2203.10428/) (2022)
- [BNS-GCN: Efficient Full-Graph Training of Graph Convolutional Networks with Partition-Parallelism and Random Boundary Node Sampling](https://mlsanity.com/paper/2203.10983/) (2022)
- [Accelerating Graph Sampling for Graph Machine Learning using GPUs](https://mlsanity.com/paper/2009.06693/) (2020)
- [Pipe-SGD: A Decentralized Pipelined SGD Framework for Distributed Deep Net Training](https://mlsanity.com/paper/1811.03619/) (2018)
- [Global Neighbor Sampling for Mixed CPU-GPU Training on Giant Graphs](https://mlsanity.com/paper/2106.06150/) (2021)
- [MariusGNN: Resource-Efficient Out-of-Core Training of Graph Neural Networks](https://mlsanity.com/paper/2202.02365/) (2022)

## Built on by (most cited first, 4 of 4)

- [A Comprehensive Benchmark on Spectral GNNs: The Impact on Efficiency, Memory, and Effectiveness](https://mlsanity.com/paper/2406.09675/) (2024)
- [MuseGNN: Forming Scalable, Convergent GNN Layers that Minimize a Sampling-Based Energy](https://mlsanity.com/paper/2310.12457/) (2023)
- [Reforge: Low-Latency Distributed GNN Serving with Selective Embedding Recomputation](https://mlsanity.com/paper/2501.08547/) (2025)
- [SNI-GNN: SmartNIC-Assisted Full-Graph GNN Training with In-Network Embedding Prediction](https://mlsanity.com/paper/2608.06441/) (2026)

## Abstract (arXiv)

A key performance bottleneck when training graph neural network (GNN) models on large, real-world graphs is loading node features onto a GPU. Due to limited GPU memory, expensive data movement is necessary to facilitate the storage of these features on alternative devices with slower access (e.g. CPU memory). Moreover, the irregularity of graph structures contributes to poor data locality which further exacerbates the problem. Consequently, existing frameworks capable of efficiently training large GNN models usually incur a significant accuracy degradation because of the currently-available shortcuts involved. To address these limitations, we instead propose FreshGNN, a general-purpose GNN mini-batch training framework that leverages a historical cache for storing and reusing GNN node embeddings instead of re-computing them through fetching raw features at every iteration. Critical to its success, the corresponding cache policy is designed, using a combination of gradient-based and staleness criteria, to selectively screen those embeddings which are relatively stable and can be cached, from those that need to be re-computed to reduce estimation errors and subsequent downstream accuracy loss. When paired with complementary system enhancements to support this selective historical cache, FreshGNN is able to accelerate the training speed on large graph datasets such as ogbn-papers100M and MAG240M by 3.4x up to 20.5x and reduce the memory access by 59%, with less than 1% influence on test accuracy.