对比主流CPU上卷积算法的能耗与延迟，找到最佳平衡点。
Benchmarking Deep Learning Convolutions on Energy-constrained CPUs
- 在ARM、Intel、AMD等多平台测试直接、GEMM、Winograd卷积方法。
- ARM Cortex-A78AE配合隐式GEMM实现102毫秒推理，功耗25.3瓦，耗能2.58焦耳。
- 首次用高精度插座级测量实现跨厂商能耗公平对比，发现寄存器估算低10%-30%。
本文评估了基于CPU的CNN推理中最新卷积算法的性能。尽管多数研究聚焦于GPU或NPU，但CPU实现仍相对未优化。我们的首个贡献是为嵌入式CPU推理提供公平的基准测试。在现代ARM、Intel、AMD和NVIDIA处理器上，评估了直接卷积、基于GEMM的卷积以及Winograd卷积，在延迟和能效方面进行了综合比较。据我们所知，这是首个使用高分辨率插座级测量平台，对跨厂商CPU功耗进行公平比较的研究。为验证方法，我们将插座级功耗测量与模型特定寄存器（MSRs）估算结果对比，发现后者低估了卷积推理功耗10%–30%。结果表明，结合隐式GEMM实现的ARM Cortex-A78AE CPU在延迟与功耗之间取得最佳平衡，实现了ResNet50v1.5的102毫秒推理，平均功耗25.3瓦，总能耗2.58焦耳。
原文摘要 · Abstract (English)
This work evaluates State-of-the-Art convolution algorithms for CPU-based CNN inference. Although most prior studies focus on GPUs or NPUs, CPU implementations remain comparatively under-optimized. Our first contribution is to provide fair benchmarking for embedded CPU inference. We evaluate direct, GEMM-based, and Winograd convolutions across modern CPUs from ARM, Intel, AMD, and NVIDIA vendors, considering both latency and energy efficiency. To the best of our knowledge, this is the first study to present a fair, cross-vendor comparison of CPU energy consumption using a high-resolution socket-level measurement platform. To validate our methodology, we further compare socket-level power measurements with estimates derived from model-specific registers (MSRs), finding that MSRs underestimate the power consumption of convolution inference by 10--30%. Our results show that the ARM\R Cortex-A78AE CPU combined with an implicit GEMM convolution implementation offers the best trade-off between latency and power consumption, achieving ResNet50v1.5 inference in 102 ms with an average power of 25.3 W, corresponding to 2.58 J.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。