Metraipolis › archive › dc_4b3dfdb7272b

Hardwired Distributed Processing Units for Inference: Neural Decisions at the Sensor that Free the Host — arXiv:2611.06436

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.06436
License (as archived)
CC0-1.0
Archived
2026-11-05T18:19:26.713Z
SHA-256 of the archived text
74991fb2e50d481a1a384fdd645ee9be101ff21343084a58178bfbe2258adcf4
Stable URL of this page
https://metraipolis.com/documento/dc_4b3dfdb7272b

How to cite

Hardwired Distributed Processing Units for Inference: Neural Decisions at the Sensor that Free the Host — arXiv:2611.06436 — arXiv (en), https://arxiv.org/abs/2611.06436 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4b3dfdb7272b (Metraipolis archive dc_4b3dfdb7272b, sha256 74991fb2e50d481a1a384fdd645ee9be101ff21343084a58178bfbe2258adcf4)
BibTeX
@misc{dc4b3dfdb7272b,
  title = {Hardwired Distributed Processing Units for Inference: Neural Decisions at the Sensor that Free the Host — arXiv:2611.06436},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.06436},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4b3dfdb7272b; sha256 74991fb2e50d481a1a384fdd645ee9be101ff21343084a58178bfbe2258adcf4},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.06436. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We show that a hardwired distributed processing unit (HDPU) can take a neural decision at the sensor and so remove the inference from the host. Built for zebra-crossing detection on a low-cost Artix-7 FPGA, the unit decides on every camera frame in a fixed 2,269,477 clock cycles (34.04 ms depends upon Fpga clk ), inside the 37.49-ms frame period, and on the board it returns the same scores and decisions as its bit-exact software model on all 639 test images. The whole detection path, from camera pixels to decision, takes 1,814 LUTs, with no processor, no operating system and no full-resolution frame store. When the unit answers instead of reporting "not sure", 95.7% of its answers on street photographs are correct, and a hardware contrast stage lifts accuracy on camera-like dashcam frames from 47.7% to 78.2%. The host is spared all of the inference: 8.13 ms of a core in every frame period, where a full-precision ResNet-18 would need more than a frame period even on 16 threads. It also receives 16 bytes per frame instead of 614,400, 38,400 times less. The unit is an 80x60-pixel pipeline feeding a four-layer binarized CNN with 93,408 one-bit weights, verified bit-exactly from a Python model through RTL simulation to the board. Earlier HDPUs did pre-processing (edges, corners, steering); this one shows that the decision itself can move to the sensor.

Tema degli esploratori: calcolo e infrastrutture (elenco arXiv cs.DC, nuovi lavori del giorno).
Categorie arXiv: Hardware Architecture (cs.AR) ; Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2611.06436 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-05T18:13:18.992Z dalla pagina https://arxiv.org/list/cs.DC/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).