Skip to content

What is the minimum subset of cuda toolkit to install for tensorflow gpu to work? #38824

Closed
@ForgottenR

Description

@ForgottenR

System information

  • OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Windows 10
  • TensorFlow installed from (source or binary): binary
  • TensorFlow version: 2.1.0
  • Python version: 3.7.7
  • Installed using virtualenv? pip? conda?: pip
  • CUDA/cuDNN version: Cuda 10.1 && cuDNN 7.6.5
  • GPU model and memory: Nvidia Geforce GTX 950M 2G

Describe the problem
Actually, there is not much of a problem. I'm just wondering what are the necessary parts for tensorflow gpu to work. I guess a subset of CUDA runtime dynamic libraries, CUPTI, and cuDNN runtime dynamic library should suffice? However, I failed to find detailed information about this in the installation guide.

Metadata

Metadata

Assignees

Labels

TF 2.1for tracking issues in 2.1 releasestaleThis label marks the issue/pr stale - to be closed automatically if no activitystat:awaiting responseStatus - Awaiting response from authorsubtype:windowsWindows Build/Installation Issuestype:build/installBuild and install issues

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions