I am unable to install CUDA 9.0.
Error msg NVIDIA Installer failed(Visual Studio Integration failed)
(Visual Studio 2015 already installed, Windows is updated and I also tried to install CUDA 9.1 but the same issue)
Unfortunately this does not work for me. I made sure to install this before trying to install the CUDA toolkit but the installer fails on VS integration. I tried uninstalling VS 2017 and using VS 2015 instead, and that didnât work either.
I download visual studio 2017 installer and i select VC++ 2015.3 Tool Set (v140) from the list and installed it but the same issue. itâs so frustrating. even NVIDA people are helping!!!
I extract the zipped file into a folder with 7zip, then I installed the setup file with administrative privileges. Up to this point, none of them worked. But when I uncheked the visual studio integration, looks like Iâve toolkit with no visual studio support :(
Thanks 4 the tips guys !
And thanks to the Nvidia for this crappy installer btw.
Alright, I have a workaround here that allows you to install CUDA 9.1 with most VS integration.
Use the VS installer to install VC++ 2015.3 Tool Set (v140)
Use 7zip to extract the local CUDA installer to a folder.
Run setup.exe from that folder
At Options, select "Custom (Advanced)"
Under CUDA, UNCHECK Visual Studio Integration
Proceed with the rest of the installation
Once that's installed, go back to the folder you extracted everything to
Go to 'CUDAVisualStudioIntegration' within that folder
Use the two msi installers to install Nsight and NVTX
Copy the files from 'extras\visual_studio_integration\MSBuildExtensions' into
'C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\Common7\IDE\VC\VCTargets\BuildCustomizations'
AND
'C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V140\BuildCustomizations'
That should get you to the point where projects will compile and run from VS.
the files in âCudaProjectVsWizardsâ look like the templates that let you start new CUDA projects and make new CUDA files within VS. It probably wouldnât take too much research to figure out where those go.
Iâm not sure where âCUDAVisualStudioIntegrationExt.dllâ and âCUDAVisualStudioIntegration.nviâ are meant to go.
Hi, after 3 days of headaches and realizing that nvidia doesnât give a damn about this, I have finally found a solution. Just disable the Visual Studio support. This was the only way, trust me, they have compatibility issues with pretty much all the VS versions.
I didnât use Visual Studios in this new Deep Learning box I built because Iâm using Python under Anaconda. Unchecking Visual Studio Integration also worked for me as well. v9.0
NVIDIA is already completely insolent, realy no user support ⊠If Microsoft is your gold partner, why you canât just make a normal installer for VisualStudio? FacepalmâŠ
You canât find any good programmers for the company - make your product opensource and all be ok :-)
I told you many times, VisualStudio is far from the best choice, make NSIGHT for QtCreator!!!
I have been tried installing cuda toolkit 9.0/9.1 with visualstudio 17 and installation has failed
i even tried installing using visualstudio 15 but it also failed. can anyone help me?
i was trying to install on Windows 10 .
Be absolutely SURE you have the VC++ 2015.3 loaded. I thought I did, but turns out I was incorrect (I assumed it was installed when I selected ALL the C++ options when I installed Visual Studio 2017). You will know this because the:
C:\Program Files (x86)\MSBuild\Microsoft.Cpp
folder will NOT be present (hint). Go into your Visual Studio 2017 Installer and MODIFY, then select the VC++ 2015.3 libraries in the âIndividual Componentsâ section. THEN follow Kooshiâs instructions. Works like a champ, and Iâm on Windows 10 Version 1709 (OS Build 16299.431).
I went around in circles trying everything to get the Cuda environment to install. My computer is win 10 pro, i7, x99 motherboard, 64GB DDR4 and 1080ti graphics card. I systematically tried not installing a single part to see if one was affecting installation. Turns out Visual Studio integration is not functioning correctly. That unit will not function with any version of Visual Studio from Community to Enterprise.
This should have been picked up in even basic testing, before deployment.
I have problem installing the cuda toolkit. I have run the installation but the devicequerry doesnât go throught:
CUDA Device Query (Runtime API) version (CUDART static linking)
cudaGetDeviceCount returned 35
â CUDA driver version is insufficient for CUDA runtime version
Result = FAIL