How to run Nvidia CUDA v8.0.27 in Visual Studio on Windows 10 ?

I am trying to render in Blender using GPU Cycles rendering. I get this message: “Cuda binary kernel for this graphics card compute capability (6.1) not found.”
My existing Nvidia Driver is version - 368.81
I have this installed - but can’t open it C:\Program Files\NVIDIA GPU Computing Toolkit]CUDA\v8.0
cuda_8.0.27_win10.exe
To install this it says I need to install
https://www.visualstudio.com/en-us
VSCodeSetup-stable.exe
from:
https://az764295.vo.msecnd.net
C:\ProgramFiles (x86)Microsoft VS Code
Visual Studio Code
installed GitHub to be able to install Python/Visual Studio
https://guides.github.com/introduction/flow/
I can’t find instructions on how to do this & what I find makes no sense - how to run Nvidia CUDA in Visual Studio (This does not list Windows 10 as a supported operating system)
http://developer.download.nvidia.com/compute/cuda/6_5/rel/docs/CUDA_Getting_Started_Windows.pdfz

If you’ve installed cuda_8.0.27_win10.exe on your machine, do a file search for

CUDA_Installation_Guide_Windows.pdf

and start reading. If you haven’t installed it, then the same file is available here:

https://developer.nvidia.com/compute/cuda/8.0/rc/docs/sidebar/CUDA_Installation_Guide_Windows-pdf

But you may need to be logged in with your developer credentials to access that link – not sure. (Incidentally, this link is found on the page where you dowloaded the CUDA 8 windows installer)

It clearly indicates win10 as a supported operating system. And yes, you must have an appropriate version of VS installed for this to be useful - the appropriate versions and system requirements are covered in the doc. Note that the VS2015 versions listed there assume VS2015 update 1 - Update 2 and 3 are not supported - yet.

Whether or not all this will help you resolve the blender issue I cannot say. It appears that whatever executable you are using has been compiled without PTX, therefore it must include a compiled binary object for your GPU. To fix that, you would have to be able to recompile the source code for that under the CUDA 8 toolkit.