关键词: 扩散模型, 模型加速, 边缘部署, 软硬件协同设计, 高效推理

Abstract: With the development of neural networks, Diffusion models have achieved remarkable success in image generation tasks due to their unique generative process. However, this performance comes at the cost of substantial computational overheads and complex network structures, posing significant challenges for efficient deployment, particularly on edge devices with limited resources. To address these bottlenecks, researchers have proposed a range of acceleration algorithms and co-designed software-hardware frameworks. This survey provides a comprehensive overview of state-of-the-art acceleration techniques for diffusion models, covering both algorithmic optimizations suitable for general-purpose computing platforms and hardware-aware designs tailored for efficient deployment.

Key words: diffusion models, model acceleration, edge deployment, hardware-software co-design, efficient inference