外观
论文arXiv 2610.0506224 天前
超越 LLM Serving：面向具身 AI 系统设计的 Vision-Language-Action 工作负载表征
原文链接：Beyond LLM Serving: Characterizing Vision-Language-Action Workloads for Embodied AI System Design（arXiv 2610.05062 · 发布日期：2026-10-04）
Beyond LLM Serving: Characterizing Vision-Language-Action Workloads for Embodied AI System Design
本文针对VLA控制周期推理期限问题，通过批1性能剖析和闭环实验，在边缘GPU服务器与机载SoC上表征工作负载。
核心要点
- VLA若推理超期会让机器人基于陈旧观测行动，降低任务成功率。
- 研究聚焦单机器人批1控制场景，分析多类VLA组件的运行行为。
- 实验表明动作张量维度、平台均衡与GPU频率会影响瓶颈和能耗-时延折中点。
- 闭环中推理与动作执行重叠带来精度、速度、能耗权衡，配置无统一Pareto最优。
验证范围与局限：验证范围为四个代表性VLA模型，在一台边缘GPU服务器和两个机载SoC上的单次推理剖析及43,200个闭环回合。
英文摘要 (Abstract)
Vision-language-action (VLA) models translate multimodal observations into low-level robot actions. During robot operation, each control period sets an inference deadline, and overruns leave the robot acting on stale observations, reducing task success. Meeting this deadline motivates on-device or nearby edge execution, where a single robot requires batch-1 inference outside the design point of LLM serving systems. Although VLA architectures combine familiar vision-language, autoregressive, and diffusion-style components, their runtime behavior in this batch-1 control setting remains uncharacterized. We characterize four representative VLA models on an edge GPU server and two onboard SoCs, using single-inference profiling and 43,200 closed-loop episodes. Action tensor dimensionality determines whether a stage is memory- or compute-bound, platform balance can shift that bottleneck, and GPU frequency scaling yields a platform-dependent energy-latency sweet spot. In closed-loop operation, overlapping inference with action execution creates an accuracy-speed-energy tradeoff, and no configuration is Pareto-dominant across deployment SLOs. These results guide joint design of VLA model architectures, hardware, and runtime policies.
作者：SeongHun Jung, Sieun Moon, Jiyoung Jeong, Jimin Lee, Jaehyuk Huh
来源
本页事实对应的公开页面，以原文为准。