More Like this

1. MATCHA: a fast and energy-efficient accelerator for fully homomorphic encryption over the torus

   https://doi.org/10.1145/3489517.3530435  

   Jiang, Lei; Lou, Qian; Joshi, Nrushad (July 2022, ACM/IEEE Design Automation Conference)

   Fully Homomorphic Encryption over the Torus (TFHE) allows arbitrary computations to happen directly on ciphertexts using homomorphic logic gates. However, each TFHE gate on state-of-the-art hardware platforms such as GPUs and FPGAs is extremely slow (> 0.2ms). Moreover, even the latest FPGA-based TFHE accelerator cannot achieve high energy efficiency, since it frequently invokes expensive double-precision floating point FFT and IFFT kernels. In this paper, we propose a fast and energy-efficient accelerator, MATCHA, to process TFHE gates. MATCHA supports aggressive bootstrapping key unrolling to accelerate TFHE gates without decryption errors by approximate multiplication-less integer FFTs and IFFTs, and a pipelined datapath. Compared to prior accelerators, MATCHA improves the TFHE gate processing throughput by 2.3x, and the throughput per Watt by 6.3x. 

   more » « less
2. Glyph: Fast and Accurately Training Deep Neural Networks on Encrypted Data

   Qian, Lou; Bo, Feng; Geoffrey, Fox; Lei, Jiang (October 2020, Advances in neural information processing systems)

   null (Ed.)

   Because of the lack of expertise, to gain benefits from their data, average users have to upload their private data to cloud servers they may not trust. Due to legal or privacy constraints, most users are willing to contribute only their encrypted data, and lack interests or resources to join deep neural network (DNN) training in cloud. To train a DNN on encrypted data in a completely non-interactive way, a recent work proposes a fully homomorphic encryption (FHE)-based technique implementing all activations by \textit{Brakerski-Gentry-Vaikuntanathan} (BGV)-based lookup tables. However, such inefficient lookup-table-based activations significantly prolong private training latency of DNNs. In this paper, we propose, Glyph, an FHE-based technique to fast and accurately train DNNs on encrypted data by switching between TFHE (Fast Fully Homomorphic Encryption over the Torus) and BGV cryptosystems. Glyph uses logic-operation-friendly TFHE to implement nonlinear activations, while adopts vectorial-arithmetic-friendly BGV to perform multiply-accumulations (MACs). Glyph further applies transfer learning on DNN training to improve test accuracy and reduce the number of MACs between ciphertext and ciphertext in convolutional layers. Our experimental results show Glyph obtains state-of-the-art accuracy, and reduces training latency by 69%~99% over prior FHE-based privacy-preserving techniques on encrypted datasets. 

   more » « less
3. OFHE: An Electro-Optical Accelerator for Discretized TFHE

   Zheng, Mengxin; Chu, Cheng; Lou, Qian; Youngblood, Nathan; Li, Mo; Moazeni, Sajjad; Jiang, Lei (May 2024, Proceedings International Symposium on Low Power Electronics and Design)

   This paper presents \textit{OFHE}, an electro-optical accelerator designed to process Discretized TFHE (DTFHE) operations, which encrypt multi-bit messages and support homomorphic multiplications, lookup table operations and full-domain functional bootstrappings. While DTFHE is more efficient and versatile than other fully homomorphic encryption schemes, it requires 32-, 64-, and 128-bit polynomial multiplications, which can be time-consuming. Existing TFHE accelerators are not easily upgradable to support DTFHE operations due to limited datapaths, a lack of datapath bit-width reconfigurability, and power inefficiencies when processing FFT and inverse FFT (IFFT) kernels. Compared to prior TFHE accelerators, OFHE addresses these challenges by improving the DTFHE operation latency by 8.7\%, the DTFHE operation throughput by $$57\%$$, and the DTFHE operation throughput per Watt by $$94\%$$. 

   more » « less
4. Encrypted Value Iteration and Temporal Difference Learning over Leveled Homomorphic Encryption

   Suh, J. & (May 2021, American Control Conference)

   We consider an architecture of confidential cloud-based control synthesis based on Homomorphic Encryption (HE). Our study is motivated by the recent surge of data-driven control such as deep reinforcement learning, whose heavy computational requirements often necessitate an outsourcing to the third party server. To achieve more flexibility than Partially Homomorphic Encryption (PHE) and less computational overhead than Fully Homomorphic Encryption (FHE), we consider a Reinforcement Learning (RL) architecture over Leveled Homomorphic Encryption (LHE). We first show that the impact of the encryption noise under the Cheon-Kim-Kim-Song (CKKS) encryption scheme on the convergence of the model-based tabular Value Iteration (VI) can be analytically bounded. We also consider secure implementations of TD(0), SARSA(0) and Z-learning algorithms over the CKKS scheme, where we numerically demonstrate that the effects of the encryption noise on these algorithms are also minimal. 

   more » « less
5. ILA: Correctness via Type Checking for Fully Homomorphic Encryption

   https://doi.org/10.1145/3719027.3765068  

   Gollamudi, Tarakaram; Gollamudi, Anitha; Gancher, Joshua (November 2025, ACM)

   Not AvailableRLWE-based Fully Homomorphic Encryption (FHE) schemes add some small noise to the message during encryption. The noise accumulates with each homomorphic operation. When the noise exceeds a critical value, the FHE circuit produces an incorrect output. This makes developing FHE applications quite subtle, as one must closely track the noise to ensure correctness. However, existing libraries and compilers offer limited support to statically track the noise. Additionally, FHE circuits are also plagued by wraparound errors that are common in finite modulus arithmetic. These two limitations of existing compilers and libraries make FHE applications too difficult to develop with confidence. In this work, we present a correctness-oriented IR, Intermediate Language for Arithmetic circuits, for type-checking circuits intended for homomorphic evaluation. Our IR is backed by a type system that tracks low-level quantitative bounds (e.g., ciphertext noise) without using the secret key. Using our type system, we identify and prove a strong functional correctness criterion for ILA circuits. Additionally, we have designed ILA to be maximally general: our core type system does not directly assume a particular FHE scheme, but instead axiomatizes a model of FHE. We instantiate this model with the exact FHE schemes (BGV, BFV and TFHE), and obtain functional correctness for free. We implement a concrete type checker ILA, parameterized by the noise estimators for three popular FHE libraries (OpenFHE, SEAL and TFHE-rs). Additionally, our novel algorithm for inferring the placement of modulus switching, a common noise management operation, uses the ILA type checker as a validation framework to ensure the correctness of the optimization. Evaluation shows that ILA type checker is sound (always detects noise overflows), practical (noise estimates are tight) and efficient. 

   more » « less