arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.00477cs.DC
收藏

我们是否需要张量核心来进行刻度计算?

Do We Need Tensor Cores for Stencil Computations?

Shanghai Jiao Tong University(上海交通大学);Shanghai Enflame Technology Co.,Ltd.(上海燧原科技有限公司)

Qiqi Gu, Chenpeng Wu, Heng Shi, Jianguo Yao, Haibing Guan

AI总结:

本文通过分析张量核心与刻度计算的性能关系,揭示了张量核心在特定刻度工作负载下的适用性,并展示了稀疏张量核心扩展加速空间的潜力。

摘要:

刻度计算是科学计算中的基石,是流体动力学、天气模拟等领域的关键内核。尽管刻度计算通常被视为内存受限,因此不适合以计算为中心的张量核心,但近期的实证研究显示,在应用张量核心后,出现了显著的加速效果,形成明显的矛盾。本文通过系统分析张量核心上的刻度计算性能,解决这一矛盾。我们首先回顾了刻度在张量核心上的适应情况,量化了为满足硬件约束而引入的计算冗余。这些指标随后被整合到一个增强的性能模型中,明确考虑了时间融合驱动的算术强度变化。基于此公式,我们推导出分析张量核心适用于不同刻度工作负载的分析标准。通过分类操作区域,我们确定了有效加速的特定“甜点”,并进一步展示了稀疏张量核心如何扩展这一有利的设计空间。在NVIDIA GPU上对SOTA实现进行的广泛评估,包括DRStencil、EBISU、ConvStencil和SPIDER,验证了我们的性能模型和分析标准。这些结果展示了我们的方法在指导刻度性能优化方面的有效性。

英文摘要:

Stencil computation constitutes a cornerstone of scientific computing, serving as a critical kernel in domains ranging from fluid dynamics to weather simulation. While stencil computations are conventionally regarded as memory-bound and thus unsuitable for compute-centric Tensor Cores, recent empirical studies have demonstrated significant speedups after applying Tensor Cores, forming an apparent contradiction. This paper resolves this contradiction by conducting a systematic performance analysis of stencil computations on Tensor Cores. We begin by revisiting the adaptation of stencils onto Tensor Cores, quantifying the computational redundancy introduced by the transformations required to satisfy hardware constraints. These metrics are subsequently integrated into an enhanced performance model that explicitly accounts for the arithmetic intensity shifts driven by temporal fusion. Guided by this formulation, we derive analytical criteria to determine the suitability of Tensor Cores for varying stencil workloads. By classifying operational regions, we identify the specific \textit{sweet spot} for effective acceleration and further demonstrate how Sparse Tensor Cores expand this profitable design space. Extensive evaluations on NVIDIA GPUs across SOTA implementations, including DRStencil, EBISU, ConvStencil, and SPIDER, validate our performance model and analytical criteria. These results demonstrate the effectiveness of our approach in guiding stencil performance optimization.

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。

↑