# Standing on the Shoulders of Giants: Hardware and Neural Architecture Co-Search with Hot Start

arXiv: https://arxiv.org/abs/2007.09087 · HTML: https://mlsanity.com/paper/2007.09087/

## Key facts

- Cited by 22 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Jul 2020.

## Built on by (most cited first, 22 of 22)

- [A Comprehensive Survey on Hardware-Aware Neural Architecture Search](https://mlsanity.com/paper/2101.09336/) (2021)
- [The Larger The Fairer? Small Neural Networks Can Achieve Fairness for Edge Devices](https://mlsanity.com/paper/2202.11317/) (2022)
- [Uncertainty Modeling of Emerging Device-based Computing-in-Memory Neural Accelerators with Application to Neural Architecture Search](https://mlsanity.com/paper/2107.06871/) (2021)
- [HAO: Hardware-aware neural Architecture Optimization for Efficient Inference](https://mlsanity.com/paper/2104.12766/) (2021)
- [Computation-efficient Deep Learning for Computer Vision: A Survey](https://mlsanity.com/paper/2308.13998/) (2023)
- [Can Noise on Qubits Be Learned in Quantum Neural Network? A Case Study on QuantumFlow](https://mlsanity.com/paper/2109.03430/) (2021)
- [Rethinking Co-design of Neural Architectures and Hardware Accelerators](https://mlsanity.com/paper/2102.08619/) (2021)
- [Auto-NBA: Efficient and Effective Search Over the Joint Space of Networks, Bitwidths, and Accelerators](https://mlsanity.com/paper/2106.06575/) (2021)
- [Quantum Neural Network Compression](https://mlsanity.com/paper/2207.01578/) (2022)
- [Exploration of Quantum Neural Architecture by Mixing Quantum Neuron Designs](https://mlsanity.com/paper/2109.03806/) (2021)
- [On-Device Unsupervised Image Segmentation](https://mlsanity.com/paper/2303.12753/) (2023)
- [Dancing along Battery: Enabling Transformer with Run-time Reconfigurability on Mobile Devices](https://mlsanity.com/paper/2102.06336/) (2021)
- [One Proxy Device Is Enough for Hardware-Aware Neural Architecture Search](https://mlsanity.com/paper/2111.01203/) (2021)
- [Special Session: Towards an Agile Design Methodology for Efficient, Reliable, and Secure ML Systems](https://mlsanity.com/paper/2204.09514/) (2022)
- [EH-DNAS: End-to-End Hardware-aware Differentiable Neural Architecture Search](https://mlsanity.com/paper/2111.12299/) (2021)
- [RoHNAS: A Neural Architecture Search Framework with Conjoint Optimization for Adversarial Robustness and Hardware Efficiency of Convolutional and Capsule Networks](https://mlsanity.com/paper/2210.05276/) (2022)
- [RBFleX-NAS: Training-Free Neural Architecture Search Using Radial Basis Function Kernel and Hyperparameter Detection](https://mlsanity.com/paper/2503.22733/) (2025)
- [How to Reach Real-Time AI on Consumer Devices? Solutions for Programmable and Custom Architectures](https://mlsanity.com/paper/2106.15021/) (2021)
- [A Semi-Decoupled Approach to Fast and Optimal Hardware-Software Co-Design of Neural Accelerators](https://mlsanity.com/paper/2203.13921/) (2022)
- [TinyML Design Contest for Life-Threatening Ventricular Arrhythmia Detection](https://mlsanity.com/paper/2305.05105/) (2023)
- [Enabling Cross-Domain Communication: How to Bridge the Gap between AI and HW Engineers](https://mlsanity.com/paper/2104.03780/) (2021)
- [BSC: Block-based Stochastic Computing to Enable Accurate and Efficient TinyML](https://mlsanity.com/paper/2111.06686/) (2021)

## Abstract (arXiv)

Hardware and neural architecture co-search that automatically generates Artificial Intelligence (AI) solutions from a given dataset is promising to promote AI democratization; however, the amount of time that is required by current co-search frameworks is in the order of hundreds of GPU hours for one target hardware. This inhibits the use of such frameworks on commodity hardware. The root cause of the low efficiency in existing co-search frameworks is the fact that they start from a "cold" state (i.e., search from scratch). In this paper, we propose a novel framework, namely HotNAS, that starts from a "hot" state based on a set of existing pre-trained models (a.k.a. model zoo) to avoid lengthy training time. As such, the search time can be reduced from 200 GPU hours to less than 3 GPU hours. In HotNAS, in addition to hardware design space and neural architecture search space, we further integrate a compression space to conduct model compressing during the co-search, which creates new opportunities to reduce latency but also brings challenges. One of the key challenges is that all of the above search spaces are coupled with each other, e.g., compression may not work without hardware design support. To tackle this issue, HotNAS builds a chain of tools to design hardware to support compression, based on which a global optimizer is developed to automatically co-search all the involved search spaces. Experiments on ImageNet dataset and Xilinx FPGA show that, within the timing constraint of 5ms, neural architectures generated by HotNAS can achieve up to 5.79% Top-1 and 3.97% Top-5 accuracy gain, compared with the existing ones.