arXiv cs.AI· Youngjin Jo, Jihwan Kim, Minseok Choi, Kwanghoon Choi, Jusung Kang, Jaein Kim·· 10 小时前AI 评分22
边缘 NPU 厂商引导优化基准测试：从纯 NPU 到端到端性能
Benchmarking Vendor-Guided Optimizations on Edge NPUs: From NPU Only to End-to-End Performance. Extended technical report
AI 导读
一项基准测试方法将边缘 NPU 性能拆分为 Baseline、Vendor Guided NPU Only（VG-NPU）和 Vendor Guided End-to-End（VG-E2E）三种配置，在四款边缘 NPU 和三个视觉负载上对比。
正文 · 原文
Abstract:Edge NPU performance depends not only on hardware capabilities, but also on model transformation and the overall deployment pipeline. We propose a benchmarking methodology that separates these effects by comparing three configurations: Baseline, Vendor Guided NPU Only (VG-NPU), and Vendor Guided End-to-End (VG-E2E), where VG-NPU preserves common model I/O and host-side processing while VG-E2E follows the vendor-native deployment pipeline. We apply this methodology to four edge NPUs and three vision workloads. VG-NPU improves both accuracy and Runtime-call latency in 10 of 12 NPU--workload combinations, whereas VG-E2E reduces System-level latency in all combinations but improves accuracy in only four. Moreover, NPU-level performance rankings do not always match End-to-End rankings, and VG-E2E provides an additional 10.4--70.7\% reduction in System-level latency over VG-NPU. These results demonstrate the importance of evaluating vendor-guided optimization separately at the controlled model-transformation and full deployment levels when comparing heterogeneous commercial Edge NPU platforms in practice.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.13563 [cs.AI] |
| (or arXiv:2611.13563v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.13563 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Kwanghoon Choi [view email]
[v1]
Tue, 10 Nov 2026 04:31:20 UTC (121 KB)
来源：arXiv cs.AI · arxiv.org