如何知道支持特定 gpu（如 gtx1080）的最低 cuda 工具包版本

Question

对于像 gtx1080 这样的特定 gpu，我想知道哪些 cuda 工具包版本支持它。浏览了nvidia的官网，没有找到具体的结果。

Answer 1

您可以先在以下站点查看 GTX1080 的计算能力 (CC)。它是 CC 6.1。

https://developer.nvidia.com/cuda-gpus

然后检查下面的CUDA doc，看看它是否在当前版本的CUDA 7.5的支持CC列表中。不是。

http://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#compute-capabilities

所以它应该在未来的版本中得到支持。目前即将发布的版本是 CUDA 8。您可以在其文档中找到它。

---

如果您不确定文档的版本，您可以找到与特定 CUDA 安装相关的文档，例如

/usr/local/cuda-7.5/doc

CUDA编译器的帮助信息也给出了支持的CC列表。

$ nvcc --help
--gpu-code <code>,...                      (-code)                           
        Specify the name of the NVIDIA GPU to assemble and optimize PTX for.
        nvcc embeds a compiled code image in the resulting executable for each specified
        <code> architecture, which is a true binary load image for each 'real' architecture
        (such as sm_20), and PTX code for the 'virtual' architecture (such as compute_20).
        During runtime, such embedded PTX code is dynamically compiled by the CUDA
        runtime system if no binary load image is found for the 'current' GPU.
        Architectures specified for options '--gpu-architecture' and '--gpu-code'
        may be 'virtual' as well as 'real', but the <code> architectures must be
        compatible with the <arch> architecture.  When the '--gpu-code' option is
        used, the value for the '--gpu-architecture' option must be a 'virtual' PTX
        architecture.
        For instance, '--gpu-architecture=compute_35' is not compatible with '--gpu-code=sm_30',
        because the earlier compilation stages will assume the availability of 'compute_35'
        features that are not present on 'sm_30'.
        Allowed values for this option:  'compute_20','compute_30','compute_32',
        'compute_35','compute_37','compute_50','compute_52','compute_53','sm_20',
        'sm_21','sm_30','sm_32','sm_35','sm_37','sm_50','sm_52','sm_53'.