# Leveraging Neural Graph Compilers in Machine Learning Research for Edge-Cloud Systems

arXiv: https://arxiv.org/abs/2504.20198 · HTML: https://mlsanity.com/paper/2504.20198/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 34 references that are in the map.
- First posted on arXiv 28 Apr 2025.

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

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks](https://mlsanity.com/paper/1905.11946/) (2019)
- [Training data-efficient image transformers & distillation through attention](https://mlsanity.com/paper/2012.12877/) (2020)
- [A ConvNet for the 2020s](https://mlsanity.com/paper/2201.03545/) (2022)
- [Knowledge Distillation: A Survey](https://mlsanity.com/paper/2006.05525/) (2020)
- [A Survey on Deep Neural Network Pruning-Taxonomy, Comparison, Analysis, and Recommendations](https://mlsanity.com/paper/2308.06767/) (2023)
- [Edge Intelligence: The Confluence of Edge Computing and Artificial Intelligence](https://mlsanity.com/paper/1909.00560/) (2019)
- [The Deep Learning Compiler: A Comprehensive Survey](https://mlsanity.com/paper/2002.03794/) (2020)
- [AI and ML Accelerator Survey and Trends](https://mlsanity.com/paper/2210.04055/) (2022)
- [Benchmarking of DL Libraries and Models on Mobile Devices](https://mlsanity.com/paper/2202.06512/) (2022)
- [FrankenSplit: Efficient Neural Feature Compression with Shallow Variational Bottleneck Injection for Mobile Edge Computing](https://mlsanity.com/paper/2302.10681/) (2023)
- [Impact of Thermal Throttling on Long-Term Visual Inference in a CPU-based Edge Device](https://mlsanity.com/paper/2010.06291/) (2020)
- [FOOL: Addressing the Downlink Bottleneck in Satellite Computing with Neural Feature Compression](https://mlsanity.com/paper/2403.16677/) (2024)
- [Deep Learning Library Testing: Definition, Methods and Challenges](https://mlsanity.com/paper/2404.17871/) (2024)
- [Adversarial Robustness of Bottleneck Injected Deep Neural Networks for Task-Oriented Communication](https://mlsanity.com/paper/2412.10265/) (2024)

## Built on by (most cited first, 2 of 2)

- [GraphNet: A Large-Scale Computational Graph Dataset for Tensor Compiler Research](https://mlsanity.com/paper/2510.24035/) (2025)
- [Clear-Weighted Bit Allocation for Satellite Downlinks](https://mlsanity.com/paper/2608.01457/) (2026)

## Abstract (arXiv)

This work presents a comprehensive evaluation of neural network graph compilers across heterogeneous hardware platforms, addressing the critical gap between theoretical optimization techniques and practical deployment scenarios. We demonstrate how vendor-specific optimizations can invalidate relative performance comparisons between architectural archetypes, with performance advantages sometimes completely reversing after compilation. Our systematic analysis reveals that graph compilers exhibit performance patterns highly dependent on both neural architecture and batch sizes. Through fine-grained block-level experimentation, we establish that vendor-specific compilers can leverage repeated patterns in simple architectures, yielding disproportionate throughput gains as model depth increases. We introduce novel metrics to quantify a compiler's ability to mitigate performance friction as batch size increases. Our methodology bridges the gap between academic research and practical deployment by incorporating compiler effects throughout the research process, providing actionable insights for practitioners navigating complex optimization landscapes across heterogeneous hardware environments.