GPU使用率 [英] GPU Utilization

问题描述

我一直在使用NVML库来获取图形和内存使用率的值 Rodinia基准套件.我观察到，在不同的频率下，同一应用程序的利用率显示出不同的值.从Wiki链接 http://en.wikipedia.org/wiki/CPU_usage 看来，确实如此没有考虑到各种停顿，例如内存，分支等.在一个时间间隔内，这种利用率究竟是多少?以及它的值如何随频率变化而变化.

I have been using NVML library to get the values of graphics and memory utilization for Rodinia benchmark suite. I observe that with different frequencies, the utilization of the same application shows different values. From the wiki link http://en.wikipedia.org/wiki/CPU_usage it seems it does not take into account the various stalls like memory, branch etc. What exactly is this utilization measuring during a time interval? And how come its value is varying with variation in frequency.

谢谢

推荐答案

利用率的定义在

The definition of the utilization rates is given in the nvml documentation, p90:

8.12 nvmlUtilization_t Struct Reference
#include <nvml.h>
Data Fields
• unsigned int gpu
Percent of time over the past second during which one or more kernels was executing on the GPU.
• unsigned int memory
Percent of time over the past second during which global (device) memory was being read or written.

如果您更改应用程序时钟，则给定工作负载的利用率可能会有所不同(我想这就是frequency的意思).

The utilization rates for a given workload will likely vary if you change the application clocks (I assume that is what you mean by frequency).

例如，如果GPU核心时钟运行速度更快，则可能会更改工作负载的处理，并且可能会花费更少的时间来完成工作负载.

For example, if the GPU core clock runs faster, then the processing of the workload may be changed, and it may take less time to complete the workload.

这篇关于GPU使用率的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！