Late Breaking Result: FPGA-Based Emulation and Fault Injection for CNN Inference Accelerators
Filip Masar, Vojtech Mrazek, Lukas Sekanina · 2025
Read Late Breaking Result: FPGA-Based Emulation and Fault Injection for CNN Inference Accelerators by Filip Masar, Vojtech Mrazek, Lukas Sekanina (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-2501-12818.txt.
Description
Filip Masar, Vojtech Mrazek, and Lukas Sekanina build a test rig for checking that AI chips keep working when faults strike, in a 2025 short paper. The chips run neural networks that sort images, and a single hardware fault can flip an answer.
Their rig runs the chip design on a reprogrammable board and injects faults on purpose. Software checks of the same kind are slow, so hardware speed matters. In their first study, they injected faults into the multiplier units of a known accelerator design running a standard image network. They measured how far accuracy fell as faults varied. The hardware rig ran the analysis about ten times faster than software.
Catalog record
| CREATOR | Filip Masar, Vojtech Mrazek, Lukas Sekanina |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 9 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-12818