FHE data processing boosted by smart in-storage acceleration approach
Rosemary: Efficient and Adaptable In-Storage Acceleration for Fully Homomorphic Encryption
Cryptography and SecurityHardware Architecture
Summary
Fully Homomorphic Encryption (FHE) lets computers work on encrypted data, but it is very slow and costly because moving data between storage and processors takes up most of the time. The authors show that current systems spend 71% to 85% of time just moving data instead of computing. They created Rosemary, a new system that processes encrypted data directly inside storage devices using special hardware and clever software tools. This approach speeds up FHE tasks like encrypted AI and database queries by over 180 times compared to standard hardware and uses much less energy.
What this means in practice
- •For cloud service providers: Enable faster and more energy-efficient encrypted computation services by integrating in-storage FHE acceleration for workloads like private database queries.
- •For ai system engineers: Deploy encrypted deep learning inference on devices with constrained memory by offloading computation inside storage, reducing data transfer overhead.
- •For enterprise storage designers: Design next-generation SSDs that include specialized FHE processing units to accelerate privacy-preserving workloads for business and government users.$Commercial implications: This paper enables building SSDs with embedded FHE engines for secure client data processing, creating new product opportunities for storage hardware vendors.
Authors
Xuan Wang, Tianqi Zhang, Gabrielle De Micheli, Yujin Nam, Keming Fan, Augusto Vega, Minxuan Zhou, Tajana Rosing
Abstract
Fully Homomorphic Encryption (FHE) enables computation on encrypted data and is promising for privacy-preserving computing. However, FHE deployment remains constrained due to massive computational, memory, and data-movement overheads. Existing acceleration strategies often overlook the storage I/O bottleneck caused by the enormous size of FHE data in large-scale applications, as most prior works focus on small benchmarks that fit in main memory. Our studies show that state-of-the-art FHE systems spend 71\% to 85\% of their execution time moving data between storage and compute units rather than performing useful computation. To address the I/O bottleneck, we present Rosemary, a heterogeneous FHE acceleration system that incorporates a novel in-storage processing (ISP) architecture with compiler-level and algorithmic optimizations. Our key innovations include 1) an adaptable hardware-aware FHE engine inside a NAND flash-based solid-state drive (SSD) for efficient and cost-effective in-storage FHE processing, 2) a Throughput-Data-Aware Program Allocator (TDAP) that partitions FHE workloads into heterogeneous programs, and 3) a Compiler-Level Ciphertext-Aware Optimizer (CLCAO) that exploits ISP-friendly FHE kernels for low-cost ISP execution. Across a diverse set of FHE workloads, including encrypted deep learning inference and cloud database queries, Rosemary achieves geometric-mean speedups of $185.7\times$ over state-of-the-art CPU/GPU implementations and $3.7\times$ over prior FHE accelerators, while improving energy efficiency by up to $3.9\times$ and Energy-Delay-Area-Product (EDAP) by up to $16.5\times$.