PrivQuant: Communication-Efficient Private Inference with Quantized Network/Protocol Co-Optimization

TL;DR

PrivQuant is a framework that combines network quantization and protocol optimization to significantly reduce communication and latency in private DNN inference using secure two-party computation.

Contribution

It introduces a joint optimization approach for quantized network inference and 2PC protocols, achieving communication efficiency and high accuracy.

Findings

Reduces communication by up to 11 times.

Achieves latency reductions of up to 8.7 times.

Outperforms prior frameworks like SiRNN, COINN, and CoPriv.

Abstract

Private deep neural network (DNN) inference based on secure two-party computation (2PC) enables secure privacy protection for both the server and the client. However, existing secure 2PC frameworks suffer from a high inference latency due to enormous communication. As the communication of both linear and non-linear DNN layers reduces with the bit widths of weight and activation, in this paper, we propose PrivQuant, a framework that jointly optimizes the 2PC-based quantized inference protocols and the network quantization algorithm, enabling communication-efficient private inference. PrivQuant proposes DNN architecture-aware optimizations for the 2PC protocols for communication-intensive quantized operators and conducts graph-level operator fusion for communication reduction. Moreover, PrivQuant also develops a communication-aware mixed precision quantization algorithm to improve inference efficiency while maintaining high accuracy. The network/protocol co-optimization enables PrivQuant to outperform prior-art 2PC frameworks. With extensive experiments, we demonstrate PrivQuant reduces communication by $11\times, 2.5\times \mathrm{and}~ 2.8\times$, which results in $8.7\times, 1.8\times ~ \mathrm{and}~ 2.4\times$ latency reduction compared with SiRNN, COINN, and CoPriv, respectively.