arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.13675cs.CRcs.AR
收藏

智能交通系统中未监控GPU的隐藏风险

Hidden Risks of Unmonitored GPUs in Intelligent Transportation Systems

Clemson University(克莱姆森大学)

Sefatun-Noor Puspa, Mashrur Chowdhury

AI总结:

本文探讨了智能交通系统中未监控GPU的潜在风险,指出未管理的GPU负载会无声地降低实时性能,强调了加强ITS安全措施的必要性。

摘要:

图形处理单元(GPUs)为许多智能交通系统(ITS)和自动驾驶应用提供动力,但仍然在安全和保安方面 largely unmonitored。本文强调GPU滥用作为关键盲点,展示了未管理的GPU负载如何无声地降级实时性能,证明了在ITS中需要更强的安全措施。

英文摘要:

Graphics processing units (GPUs) power many intelligent transportation systems (ITS) and automated driving applications, but remain largely unmonitored for safety and security. This article highlights GPU misuse as a critical blind spot, showing how unmanaged GPU workloads silently degrade real-time performance, demonstrating the need for stronger security measures in ITS.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑