用OpenCL记忆()缓存GPU缓冲区的最快方法是什么? [英] What is the fastest way to memset() a GPU buffer with OpenCL?

问题描述

我正在使用OpenCL，我需要memset()在全局设备内存中放置一些数组. CUDA具有类似memset()的API函数，但OpenCL没有.我阅读了此，在其中我找到了两种可能的选择:

I'm using OpenCL, and I need to memset() some array in global device memory. CUDA has a memset()-like API function, but OpenCL does not. I read this, where I found two possible alternatives:

1. 在主机上使用memset()并使用一些暂存缓冲区，然后clEnqueueWriteBuffer()将其复制到设备上的缓冲区.

2. 排队以下内核:

1. using memset() on the host with some scratch buffer, then clEnqueueWriteBuffer() to copy that to the buffer on the device.

2. Enqueueing the following kernel:

__kernel void memset_uint4(
    __global  uint4* mem,
    __private uint4  val) 
{
    mem[get_global_id(0)] = val; 
}

哪个更好?或者更确切地说，在哪种情况下/哪种平台比另一种更好?

Which is better? Or rather, under which circumstances/for which platforms is one better than the other?

注意:如果零记忆的特殊情况值得特殊对待，那也很好.

Note: If the special case of zero'ing memory merits special treatment, that would be nice to know too.

推荐答案

您可以从OpenCL v1.2使用clEnqueueFillBuffer().那正是您所需要的.而且在如何使用模式填充缓冲区方面非常灵活.

You can use clEnqueueFillBuffer() from OpenCL v1.2. That is exactly what you need. And it is very flexible on how to fill the buffer with patterns.

这是文档页面:

http://www.khronos.org/registry/cl /sdk/1.2/docs/man/xhtml/clEnqueueFillBuffer.html

如果您使用的是1.1或更低版本....那么您应该使用其他方法.

If you are on 1.1 or below.... then you should recur to other approaches.

这篇关于用OpenCL记忆()缓存GPU缓冲区的最快方法是什么?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！