Side-Channel Extraction of Dataflow AI Accelerator Hardware Parameters

TL;DR

This paper presents a lightweight side-channel attack methodology that efficiently recovers dataflow accelerator hardware parameters with high accuracy, significantly reducing attack time compared to existing methods.

Contribution

It introduces an unsupervised dimensionality reduction approach enabling fast, accurate hardware parameter recovery with minimal traces, improving attack efficiency over prior techniques.

Findings

Achieves over 95% accuracy in hardware parameter recovery

Reduces attack time by up to 940 times compared to state-of-the-art

Requires only 4 traces on average for successful attack

Abstract

Dataflow neural network accelerators efficiently process AI tasks on FPGAs, with deployment simplified by ready-to-use frameworks and pre-trained models. However, this convenience makes them vulnerable to malicious actors seeking to reverse engineer valuable Intellectual Property (IP) through Side-Channel Attacks (SCA). This paper proposes a methodology to recover the hardware configuration of dataflow accelerators generated with the FINN framework. Through unsupervised dimensionality reduction, we reduce the computational overhead compared to the state-of-the-art, enabling lightweight classifiers to recover both folding and quantization parameters. We demonstrate an attack phase requiring only 337 ms to recover the hardware parameters with an accuracy of more than 95% and 421 ms to fully recover these parameters with an averaging of 4 traces for a FINN-based accelerator running a CNN, both using a random forest classifier on side-channel traces, even with the accelerator dataflow fully loaded. This approach offers a more realistic attack scenario than existing methods, and compared to SoA attacks based on tsfresh, our method requires 940x and 110x less time for preparation and attack phases, respectively, and gives better results even without averaging traces.