Metraipolis › archive › dc_bb8f874e8bb1
RAMP: Robust Adaptive Mixed-Precision Quantization for Edge CPU Vision Models — arXiv:2609.28262
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.28262
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T10:07:29.446Z
- SHA-256 of the archived text
01e7acd0820b5b0ab2541fb0f614fe0a5c192add754d231907ee6fe56d540442- Stable URL of this page
https://metraipolis.com/documento/dc_bb8f874e8bb1
How to cite
RAMP: Robust Adaptive Mixed-Precision Quantization for Edge CPU Vision Models — arXiv:2609.28262 — arXiv (en), https://arxiv.org/abs/2609.28262 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bb8f874e8bb1 (Metraipolis archive dc_bb8f874e8bb1, sha256 01e7acd0820b5b0ab2541fb0f614fe0a5c192add754d231907ee6fe56d540442)
BibTeX
@misc{dcbb8f874e8bb1,
title = {RAMP: Robust Adaptive Mixed-Precision Quantization for Edge CPU Vision Models — arXiv:2609.28262},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.28262},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bb8f874e8bb1; sha256 01e7acd0820b5b0ab2541fb0f614fe0a5c192add754d231907ee6fe56d540442},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.28262. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Deploying deep learning models on edge CPUs is bottlenecked by computational and memory constraints. Mixed-precision quantization promises to reduce inference latency while preserving accuracy. However, quantization affects different layer types in inconsistent ways, so identifying where accuracy loss is minimized and latency reduction is maximized is critical, as the effect accumulates over a full deployment into substantial savings or unacceptable task degradation. Such identification relies on sensitivity metrics, proxies that estimate layer-wise degradation without evaluating the task accuracy of every candidate policy. Nevertheless, widely used metrics fail systematically on modern architectures. We present a systematic empirical study of 13 sensitivity metrics for layer-wise INT8 quantization across four distinctly different neural networks, and validate the resulting policies on two ARM64 platforms. Gradient-based sensitivity methods fail on 4 out of 8 model-hardware configurations and weight-based statistics on 2. In contrast, the Jensen-Shannon Divergence achieves zero catastrophic failures, reliably isolating the layers that cannot be safely quantized. A sensitivity metric alone does not define a policy, and the fixed thresholds typically used for that step are fragile over the highly skewed distributions of modern architectures. We address this with K-Means clustering, achieving near-lossless accuracy and a mean speed-up of $1.81\times$ over the full-precision model. Finally, we reveal that excluding from quantization the layers whose speed-up is negligible, regardless of their sensitivity, can be counterproductive, as it induces computational graph fragmentation and disables operator fusion. Our results yield concrete allocation policies for practitioners and researchers deploying quantized vision models on heterogeneous edge CPUs, without GPU access or gradient computation. Tema degli esploratori: calcolo e infrastrutture (retro-catalogo arXiv: annunci del 2026-09-24 in cs.PF). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Machine Learning (cs.LG); Performance (cs.PF). Pagina del lavoro: https://arxiv.org/abs/2609.28262 — 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 2026-10-08T09:14:54.778Z dalla pagina https://arxiv.org/catchup/cs.PF/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).