Abstract:Homomorphic encryption (HE) allows direct computations on encrypted data. Despite numerous research efforts, the practicality of HE schemes remains to be demonstrated. In this regard, the enormous size of ciphertexts involved in HE computations degrades computational efficiency. Near-memory processing (NMP) and computing-in-memory (CiM)—paradigms where computation is done within the memory boundaries—represent architectural solutions for reducing latency and energy associated with data transfers in data-intensive applications, such as HE. This article introduces CiM-HE, a CiM architecture that can support operations for the Brakerski/Fan–Vercauteren (B/FV) scheme, a somewhat HE scheme for general computation. CiM-HE hardware consists of customized peripherals, such as sense amplifiers, adders, bit shifters, and sequencing circuits. The peripherals are based on CMOS technology and could support computations with memory cells of different technologies. Circuit-level simulations are used to evaluate our CiM-HE framework assuming a 6T-SRAM memory. We compare our CiM-HE implementation against: 1) two optimized CPU HE implementations and 2) a field-programmable gate array (FPGA)-based HE accelerator implementation. Compared with a CPU solution, CiM-HE obtains speedups between $4.6times $ and $9.1times $ and energy savings between $266.4times $ and $532.8times $ for homomorphic multiplications (the most expensive HE operation). Also, a set of four end-to-end tasks, i.e., mean, variance, linear regression, and inference, are up to $1.1times $ , $7.7times $ , $7.1times $ , and $7.5times $ faster (and $301.1times $ , $404.6times $ , $532.3times $ , and $532.8times $ more energy efficient). Compared with CPU-based HE in previous work, CiM-HE obtains $14.3times $ speedup and $> 2600times $ energy savings. Finally, our design offers $2.2times $ speedup with $88.1times $ energy savings compared with a state-of-the-art FPGA-based accelerator.

First Demonstration of Homomorphic Encryption using Multi-Functional RRAM Arrays with a Novel Noise-Modulation Scheme

Computing-in-Memory for Performance and Energy-Efficient Homomorphic Encryption

A Multi-Layer Parallel Hardware Architecture for Homomorphic Computation in Machine Learning

Security Enhancement for RRAM Computing System through Obfuscating Crossbar Row Connections

Stochastic switching and analog-state programmable memristor and its utilization for homomorphic encryption hardware

HYPERLOCK: In-Memory Hyperdimensional Encryption in Memristor Crossbar Array

GPU Accelerated Full Homomorphic Encryption Cryptosystem, Library and Applications for IoT Systems

1-HEMT-1-Memristor With Hardware Encryptor for Privacy-Preserving Image Processing

An Edram-Based In-Situ-Computing Processor for Homomorphic Encryption Evaluation on the Edge

A Novel DES Encryption Circuit Based on RRAM XOR Gates

Experimental Demonstration of High-order In-memory Computing Based on IGZO Charge Trapping RAM Array for Polynomial Regression Acceleration

CARM: CUDA-Accelerated RNS Multiplication in Word-Wise Homomorphic Encryption Schemes for Internet of Things

Phantom: A CUDA-Accelerated Word-Wise Homomorphic Encryption Library

Hdc-Im: Hyperdimensional Computing In-Memory Architecture Based On Rram

Intelligent Computing with RRAM

Secure and Efficient General Matrix Multiplication On Cloud Using Homomorphic Encryption

A 3D Memristor Architecture for In-Memory Computing Demonstrated with SHA3

Reconfigurable multi-core array architecture and mapping method for RNS-based homomophic encryption

Technological Exploration of Rram Crossbar Array for Matrix-Vector Multiplication

A RRAM-based Data Hiding Technique Utilizing the Impact of Form Condition on SET Performance

Implementation of Discrete Fourier Transform Using RRAM Arrays with Quasi-Analog Mapping for High-Fidelity Medical Image Reconstruction