REDACTOR: eFPGA Redaction for DNN Accelerator Security

Yazan Baddour, Ava Hedayatipour, Amin Rezaei · 2025

Read REDACTOR: eFPGA Redaction for DNN Accelerator Security by Yazan Baddour, Ava Hedayatipour, Amin Rezaei (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-18740.txt.

Description

Yazan Baddour and co-authors present REDACTOR in a 2025 paper on protecting neural network hardware. Chips that accelerate deep learning hold valuable trained designs, and thieves who copy the hardware steal that value. Embedding part of the design in a reconfigurable eFPGA hides it until an authorized bitstream unlocks it after fabrication.

The team redacts a full accelerator, choosing which modules to hide and testing both regular and fracturable lookup tables. They carry the design through synthesis, timing checks, and physical layout. Power, area, and delay overheads came out reasonable next to the security gained. The paper covers the flow from specification to silicon.

Catalog record

CREATORYazan Baddour, Ava Hedayatipour, Amin Rezaei
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE28 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-18740

Browse the full library · About the Tower of Babel