# Play It Cool: Dynamic Shifting Prevents Thermal Throttling

arXiv: https://arxiv.org/abs/2206.10849 · HTML: https://mlsanity.com/paper/2206.10849/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 6 of its 13 references that are in the map.
- First posted on arXiv 22 Jun 2022.

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

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding](https://mlsanity.com/paper/1804.07461/) (2018)
- [Dynamic Neural Networks: A Survey](https://mlsanity.com/paper/2102.04906/) (2021)
- [Slimmable Neural Networks](https://mlsanity.com/paper/1812.08928/) (2018)
- [DynaBERT: Dynamic BERT with Adaptive Width and Depth](https://mlsanity.com/paper/2004.04037/) (2020)
- [Impact of Thermal Throttling on Long-Term Visual Inference in a CPU-based Edge Device](https://mlsanity.com/paper/2010.06291/) (2020)

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

- [Enhancing User Experience in On-Device Machine Learning with Gated Compression Layers](https://mlsanity.com/paper/2405.01739/) (2024)
- [T3C: Test-Time Tensor Compression with Consistency Guarantees](https://mlsanity.com/paper/2601.01299/) (2026)

## Abstract (arXiv)

Machine learning (ML) has entered the mobile era where an enormous number of ML models are deployed on edge devices. However, running common ML models on edge devices continuously may generate excessive heat from the computation, forcing the device to "slow down" to prevent overheating, a phenomenon called thermal throttling. This paper studies the impact of thermal throttling on mobile phones: when it occurs, the CPU clock frequency is reduced, and the model inference latency may increase dramatically. This unpleasant inconsistent behavior has a substantial negative effect on user experience, but it has been overlooked for a long time. To counter thermal throttling, we propose to utilize dynamic networks with shared weights and dynamically shift between large and small ML models seamlessly according to their thermal profile, i.e., shifting to a small model when the system is about to throttle. With the proposed dynamic shifting, the application runs consistently without experiencing CPU clock frequency degradation and latency increase. In addition, we also study the resulting accuracy when dynamic shifting is deployed and show that our approach provides a reasonable trade-off between model latency and model accuracy.