Metraipolis › archive › dc_83ba1fc685ef

Efficient Block-Layer Parallel Inference for Vision-Language-Action on Hybrid Architectures — arXiv:2608.14586

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.14586
License (as archived)
CC0-1.0
Archived
2026-10-18T17:47:46.557Z
SHA-256 of the archived text
e7be5071960788d2c4d60918f2cf02cb18268163a4188fb257906f77e45952b2
Stable URL of this page
https://metraipolis.com/documento/dc_83ba1fc685ef

How to cite

Efficient Block-Layer Parallel Inference for Vision-Language-Action on Hybrid Architectures — arXiv:2608.14586 — arXiv (en), https://arxiv.org/abs/2608.14586 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_83ba1fc685ef (Metraipolis archive dc_83ba1fc685ef, sha256 e7be5071960788d2c4d60918f2cf02cb18268163a4188fb257906f77e45952b2)
BibTeX
@misc{dc83ba1fc685ef,
  title = {Efficient Block-Layer Parallel Inference for Vision-Language-Action on Hybrid Architectures — arXiv:2608.14586},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.14586},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_83ba1fc685ef; sha256 e7be5071960788d2c4d60918f2cf02cb18268163a4188fb257906f77e45952b2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.14586. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Vision-Language-Action (VLA) models are becoming a promising paradigm for autonomous driving, but their deployment on existing vehicle platforms remains difficult because they introduce both high inference latency and strong GPU-side resource pressure. In a full autonomous driving stack, this problem is even more pronounced: legacy vehicle platforms were provisioned for modular pipelines, yet after several planning-related functions are absorbed into a unified VLA model, part of the original CPU budget becomes underutilized, while the visual encoder and the main reasoning path still concentrate most computation and memory demand on the GPU. As a result, directly deploying VLA together with the rest of the onboard system can be hard under realistic GPU memory constraints. To address this issue, we present a hybrid CPU--GPU inference framework with flexible resource scheduling for autonomous driving. Our design partitions the VLA backbone at the block-layer granularity, executes the visual encoder and LLM prefix on the GPU, and offloads the LLM suffix to the CPU through a cross-frame asynchronous pipeline, thereby exposing a schedulable boundary for redistributing compute and memory pressure across heterogeneous processors. We evaluate the proposed framework on two representative driving VLA models, Orion and MindDrive. On Bench2Drive, our method reduces average latency from 521ms to 408.0ms for Orion and from 443ms to 306.2ms for MindDrive, corresponding to 21.7% and 30.9% reduction, respectively. For Orion, the estimated peak GPU memory is further reduced from 45GB to 29GB. In real-vehicle deployment under coexistence with this http URL , native Orion cannot run because the onboard GPU memory budget is insufficient, whereas the hybrid version runs successfully together with the full vehicle stack.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-18 in cs.AI).
Categorie arXiv: Distributed, Parallel, and Cluster Computing (cs.DC) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.14586 — 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 2026-10-18T16:29:13.904Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).