arXiv cs.AI· Haozhe Wang, Lina Chen, Fanqi Wan, Qi Han, Yinmin Zhang, Zheng Ge, Daxin Jiang, Junchi Yan·· 3 小时前AI 评分43
Metal Sandbox：面向 Apple Silicon 上 GPU 编程智能体的隔离执行与命令调度
Metal Sandbox: Isolated Execution and Command Scheduling for GPU Coding Agents on Apple Silicon
AI 导读
研究者发布 Metal Sandbox，一套面向 Apple Silicon 上生成 Metal kernel 的编程智能体的基础设施，提供受限文件访问、原生命令执行与跨 Mac 的命令级调度，并集成 Harbor 支持任务执行、自动验证与 rollout 收集。
正文 · 原文
Abstract:On-device inference requires kernels that make effective use of local accelerators. Efficient training and evaluation of models that generate Apple Metal kernels benefit from native macOS tools and concurrent agent sessions. These sessions alternate between model requests, development commands, and GPU execution, leaving GPU capacity idle for much of a rollout. We present Metal Sandbox, infrastructure that combines restricted per-instance file access, native command execution, and command-level scheduling across interconnected Macs. Each agent retains its workspace and private development environment. Parallel commands can run concurrently, while serial commands share one execution slot per node; eligible waiting commands can migrate to available Macs and return results to the original agent. Integration with Harbor supports task execution, automatic verification, and rollout collection through a common workflow. Controlled experiments on four Mac mini M4 nodes achieve $1.68\times$--$2.05\times$ gains in command throughput under concentrated demand. Sixteen static agent traces give an estimated $2.14\times$ sustained-throughput gain for four sessions per node versus one, assuming continuously replenished sessions and unchanged mean durations. Replaying these measured traces under dynamic scheduling reduces mean tool waiting by 18.8% and batch elapsed time by 22.3%, assuming unchanged phase durations. Metal Sandbox and its Harbor integration are open-source on GitHub.
| Comments: | 16 pages, 7 figures, 5 tables. Code available at this https URL and this https URL |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Distributed, Parallel, and Cluster Computing (cs.DC) |
| Cite as: | arXiv:2611.17294 [cs.LG] |
| (or arXiv:2611.17294v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2611.17294 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Haozhe Wang [view email]
[v1]
Thu, 12 Nov 2026 18:54:42 UTC (590 KB)
来源：arXiv cs.AI · arxiv.org