问题描述
我是 CUDA 编程的新手，我正在解决一个需要在一台机器上安装多个 GPU 的问题.我知道为了更好地进行图形编程，需要通过 SLI 组合多个 GPU.但是，对于 CUDA 编程，我是否还需要通过 SLI 组合 GPU?
I am new to CUDA programming, and I am working on a problem that requires multiple GPUs in one machine. I understand that for better graphics programming multiple GPUs need to be combined via SLI. However, for CUDA programming do I need to combine GPUs via SLI as well?
推荐答案
不，一般来说，如果您打算将 GPU 用于计算而不是纯图形应用程序，则通常不希望使用 SLI.您将能够从 CUDA 程序中将两个 GPU 作为独立设备访问.请注意，您需要在 GPU 之间明确划分工作.
No, in general you don't want to use SLI if you plan on using the GPUs for compute instead of pure graphics applications. You will be able to access both GPUs as discrete devices from within your CUDA program. Note that you will need to explicitly divide work between the GPUs.
我没有解释为什么 SLI 不适用于计算应用程序，但这是我在 Nvidia 论坛上读到的，并在 IRC 频道上从其他人那里听到的.
I don't have an explanation for why SLI isn't desirable for compute applications, but it's what I've read on the Nvidia forums and heard from others in IRC channels.
这篇关于适用于多个 GPU 的 SLI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！