Data-Rate-Aware High-Speed CNN Inference on FPGAs

Tobias Habermann, Martin Kumm · 2026

Read Data-Rate-Aware High-Speed CNN Inference on FPGAs by Tobias Habermann, Martin Kumm (2026) 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-2603-08726.txt.

Description

Tobias Habermann and Martin Kumm tune FPGA neural accelerators to the data rate each layer actually delivers. Pooling and strided convolutions shrink the stream, leaving fully unrolled units idle downstream. Their data-rate-aware design searches the design space for setups that keep every unit busy on a continuous flow, cutting arithmetic resources well below prior designs.

Catalog record

CREATORTobias Habermann, Martin Kumm
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE18667
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.08726

Browse the full library · About the Tower of Babel