AI 中文总结
研究针对现代工作负载资源利用和瓶颈的阶段级分析，提出活动图可视化技术，能综合多种指标展示性能，通过低秩GEMM和曼巴两个案例研究，揭示了仅靠传统工具难以发现的优化机会。
AI 中文摘要
我们提出了活动图，这是一种用于现代工作负载中资源利用和瓶颈的阶段级分析的可视化技术。现有工具存在权衡：屋顶线将工作负载汇总为单个点并失去所有时间概念，而剖析器和跟踪则暴露细粒度事件但模糊了性能界限。相反，活动图在单个图中描绘计算吞吐量和内存带宽利用率、计算和内存流量以及延迟。由于它们可以从分析模型、模拟或剖析数据生成，活动图捕获了理想界限和内核的实际性能。我们在两个案例研究中展示了它们：一个低秩通用矩阵乘法（GEMM），揭示了降低操作强度可以提高端到端性能这一违反直觉的结果；以及曼巴，展示了跨阶段的融合和流水线机会。在这两种情况下，我们的可视化技术都揭示了仅靠屋顶线或剖析器难以识别的优化机会。
英文摘要
We present campaign diagrams, a visualization technique for phase-level analysis of resource utilization and bottlenecks in modern workloads. Existing tools have a trade-off: rooflines aggregate a workload into a single point and lose all notion of time, while profilers and traces expose fine-grained events but obscure what bounds performance. Instead, a campaign diagram depicts compute throughput and memory bandwidth utilization, compute and memory traffic volume, and latency in a single figure. Since they can be generated from analytical models, simulations, or profiling data, campaign diagrams capture both ideal bounds and a kernel's achieved performance. We demonstrate them on two case studies: a low-rank GEMM, where they reveal the counterintuitive result that reducing operational intensity can improve end-to-end performance, and Mamba, where they expose fusion and pipelining opportunities across phases. In both cases, our visualization technique reveals optimization opportunities that are difficult to identify with rooflines or profilers alone.
Comments12 pages, 13 figures
