一句话：Cinnamon 通过新算法、编译器和硬件技术实现 FHE 加密 AI 的横向扩展，突破单体芯片限制。
全同态加密（FHE）支持密文计算但性能开销大，现有 FHE 架构多为单体芯片且仅针对小型 ML 工作负载。Cinnamon 是加速 FHE 加密的先进 ML 工作负载的框架，利用程序各层级并行性，结合新颖算法、编译器和硬件技术，构建横向扩展（scale-out）的 FHE 设计而非单体芯片。评估显示，在小程序上相较先前最佳设计性能提升 2.3×；首次展示 BERT 等大型语言模型在 FHE 下的可扩展推理，相较 CPU 加速 36,600×，推理时间从 17 小时降至 1.67 秒；其并行化策略与架构扩展降低单芯片资源需求，相比最佳单体与 chiplet 架构分别实现 5× 和 2.68× 的性价比提升。
Fully homomorphic encryption (FHE) is a promising cryptographic solution that enables computation on encrypted data, but its adoption remains a challenge due to steep performance overheads. Although recent FHE architectures have made valiant efforts to narrow the performance gap, they not only have massive monolithic chip designs but also only target small ML workloads. We present Cinnamon, a framework for accelerating state-of-the-art ML workloads that are encrypted using FHE. Cinnamon accelerates encrypted computing by exploiting parallelism at all levels of a program, using novel algorithms, compilers, and hardware techniques to create a scale-out design for FHE as opposed to a monolithic chip design. Our evaluation of the Cinnamon framework on small programs shows a 2.3× improvement in performance compared to prior state-of-the-art designs. Further, we use Cinnamon to show for the first time the scalability of large ML models such as the BERT language model in FHE. Cinnamon achieves a speedup of 36,600× compared to a CPU bringing down the inference time from 17 hours to 1.67 seconds thereby enabling new opportunities for privacy-preserving machine learning. Finally, Cinnamon's parallelization strategies and architectural extensions reduce the required resources per-chip leading to a 5× and 2.68× improvement in performance-per-dollar compared to state-of-the-art monolithic and chiplet architectures respectively.
@inproceedings{JayashankarCTZ025,
title = {Cinnamon: A Framework for Scale-Out Encrypted AI},
author = {Siddharth Jayashankar and Edward Chen and Tom Tang and Wenting Zheng and Dimitrios Skarlatos},
booktitle = {ASPLOS},
year = {2025},
doi = {10.1145/3669940.3707260}
}