性能分析方法、装置、系统、计算设备及存储介质
By analyzing kernel function performance data using the roofline model, the problem of accurately locating the performance bottleneck of graphics processor kernel functions was solved, enabling efficient performance optimization guidance.
CN116701143BActive Publication Date: 2026-07-17MOORE THREADS TECH CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- MOORE THREADS TECH CO LTD
- Filing Date
- 2023-06-07
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Existing technologies make it difficult to accurately identify the performance bottlenecks of graphics processor kernel functions, making targeted optimization impossible.
Method used
By analyzing the roofline model, kernel function performance data is obtained, computational intensity and the maximum actual computing power are determined, and the performance bottleneck area in the roofline model is used to indicate whether the target kernel function needs to be optimized, and the optimization method is selected according to the location.
Benefits of technology
It enables accurate analysis of kernel function performance, intuitively indicates performance bottlenecks, provides optimization directions, and improves the accuracy and efficiency of performance optimization.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN116701143B_ABST
Abstract
本公开涉及一种性能分析方法、装置、系统、计算设备及存储介质，所述方法包括：获取图形处理器运行目标核函数时的核函数性能数据；根据核函数性能数据确定图形处理器运行目标核函数时的计算强度和实际算力最大值；根据计算强度和实际算力最大值，在屋顶线模型中增加表示图形处理器运行目标核函数的性能的点，在该点的位置落在屋顶线模型中的性能瓶颈区域时，指示该目标核函数存在性能瓶颈，将该目标核函数设置为待优化核函数。本公开实施例的性能分析方法，借助屋顶线模型分析确定目标核函数的性能是否出现瓶颈，在保证性能分析结果的准确度的同时，便于根据性能分析结果确定性能优化方式。
Need to check novelty before this filing date? Find Prior Art