The Founder’s Guide:

How to Install the NVIDIA CUDA Driver, Toolkit, cuDNN, and TensorRT on Windows 10

The condensed tutorial with copy and paste code and screenshots

Image by Nana Dua

“The expanded version of this article uses concise explanations to help you learn what’s happening 💡”

Join the NVIDIA Developer Program:

1. Visit the official website.
2. Click “Join Now”
3. Enter email address
4. Click “Next”
5. Click “Create Account”
6. Enter user information
7. Click “Create Account”
8. Verify email address
9. Click “Submit”
10. Enter user information
11. Click “Submit”

Download Visual Studio 2019:

1. Visit the official website
2. Click “Community” next to Visual Studio 2019
3. Click “Download Visual Studio”

Install Visual Studio 2019:

1. Open the downloaded file
2. Click “Continue”
3. Check “Desktop Development with C++”
4. Click “Install”

Download the CUDA Driver:

1. Visit the official website
2. Enter graphics card information
3. Select “Studio Driver”
4. Click “Search”
5. Click “Download”
6. Click “Download”

Install the CUDA Driver:

1. Open the downloaded file
2. Click “OK”
3. Select “NVIDIA Graphics Driver”
4. Click “Agree and Continue”
5. Click “Next”
6. Click “Close”

Download the CUDA Toolkit:

1. Visit the official website: [10.0][10.1][10.2][11.0][11.1][11.2]
2. Press “Enter”
3. Click “Windows”
4. Click “x86_64”
5. Click “10”
6. Click “Exe (Local)”
7. Click “Download”

Install the CUDA Toolkit:

1. Open the downloaded file
2. Click “OK”
3. Click “Agree and Continue”
4. Click “Custom (Advanced)”
5. Click “Next”
6. Uncheck “NVIDIA GeForce Experience Components”
7. Uncheck “Driver Components”
8. Uncheck “Other Components”
9. Click “Next”
10. Click “Next”
11. Click “Close”

Download the cuDNN Libraries:

1. Visit the official website
2. Click “I Agree To the Terms of the cuDNN Software License Agreement”
3. Click “Download cuDNN v8.0.5 (November 9th, 2020), for CUDA 11.1”
4. Click “cuDNN Library for Windows (x86)”

Open PowerShell:

1. Press “⊞ Windows”
2. Enter “PowerShell” into the search bar
3. Click “Run as Administrator”

Open the Downloads Directory:

1. Copy the command from below these instructions
2. Paste the command into PowerShell
3. Press “Enter”

cd $HOME\downloads

Unzip the cuDNN Libraries:

1. Copy the command from below these instructions
2. Paste the command into PowerShell
3. Press “Enter”

expand-archive cudnn-11.1-windows-x64-v8.0.5.39.zip -destinationpath cudnn-11.1

Copy the DLL Files:

1. Copy the command from below these instructions
2. Paste the command into PowerShell
3. Press “Enter”

cp cudnn-11.1\cuda\bin\cudnn*.dll "c:\program files\nvidia gpu computing toolkit\cuda\v11.1\bin"

Copy the H Files:

1. Copy the command from below these instructions
2. Paste the command into PowerShell
3. Press “Enter”

cp cudnn-11.1\cuda\include\cudnn*.h "c:\program files\nvidia gpu computing toolkit\cuda\v11.1\include"

Copy the Lib Files:

1. Copy the command from below these instructions
2. Paste the command into PowerShell
3. Press “Enter”

cp cudnn-11.1\cuda\lib\x64\cudnn*.lib "c:\program files\nvidia gpu computing toolkit\cuda\v11.1\lib\x64"

Download TensorRT:

1. Visit the official website
2. Click “TensorRT 7”
3. Complete the survey
4. Check “I Agree To the Terms of the NVIDIA TensorRT License Agreement”
5. Click “TensorRT 7.2.2”
6. Click “TensorRT 7.2.2 for Windows10 and CUDA 11.1 & 11.2 ZIP package”

Unzip TensorRT:

1. Copy the command from below these instructions
2. Paste the command into PowerShell
3. Press “Enter”

expand-archive tensorrt-7.2.2.3.windows10.x86_64.cuda-11.1.cudnn8.0.zip -destinationpath tensorrt

Copy the TensorRT Libraries:

1. Copy the command from below these instructions
2. Paste the command into PowerShell
3. Press “Enter”

cp tensorrt\tensorrt-7.2.2.3\lib\*.dll "c:\program files\nvidia gpu computing toolkit\cuda\v11.1\bin"

Install Graph Surgeon:

1. Copy the command from below these instructions
2. Paste the command into PowerShell
3. Press “Enter”

python -m pip install tensorrt\tensorrt-7.2.2.3\graphsurgeon\graphsurgeon-0.4.5-py2.py3-none-any.whl

Install Universal Framework Format:

1. Copy the command from below these instructions
2. Paste the command into PowerShell
3. Press “Enter”

python -m pip install tensorrt\tensorrt-7.2.2.3\uff\uff-0.6.9-py2.py3-none-any.whl

Install ONNX GraphSurgeon:

1. Copy the command from below these instructions
2. Paste the command into PowerShell
3. Press “Enter”

python -m pip install tensorrt\tensorrt-7.2.2.3\onnx_graphsurgeon\onnx_graphsurgeon-0.2.6-py2.py3-none-any.whl

Open the Visual Studio File:

1. Copy the command from below these instructions
2. Paste the command into PowerShell
3. Press “Enter”

start-process “c:\programdata\nvidia corporation\cuda samples\v11.1\4_finance\blackscholes\BlackScholes_vs2019.sln”

Edit the Linker Input Properties:

1. Click the “Project” menu
2. Click “Properties”
3. Double-click “Linker”
4. Click “Input”
5. Click “Additional Dependencies”
6. Click the down arrow
7. Click “Edit”

Add the cuDNN Library:

1. Copy the library name from below these instructions
2. Paste the library name into Visual Studio 2019
3. Click “OK”
4. Click “OK”

cudnn.lib

Build the Sample:

1. Click the “Build” menu
2. Click “Build Solution”

Run the Sample:

1. Copy the command from below these instructions
2. Paste the command into PowerShell
3. Press “Enter”

cmd /k "c:\programdata\nvidia corporation\cuda samples\v11.1\bin\win64\debug\blackscholes.exe"

“Hopefully, this article helped you get the 👯‍♀️🏆👯‍♀️, remember to subscribe to get more content 🏅”

Next Steps:

This article is part of a mini-series that helps readers set up everything they need to start learning about artificial intelligence, machine learning, deep learning, and or data science. It includes articles that contain instructions with copy and paste code and screenshots to help readers get the outcome as soon as possible. It also includes articles that contain instructions with explanations and screenshots to help readers learn about what’s happening.

Linux:
01. Install and Manage Multiple Python Versions
02. Install the NVIDIA CUDA Driver, Toolkit, cuDNN, and TensorRT
03. Install the Jupyter Notebook Server
04. Install Virtual Environments in Jupyter Notebook
05. Install the Python Environment for AI and Machine Learning
WSL2:
01. Install Windows Subsystem for Linux 2
02. Install and Manage Multiple Python Versions
03. Install the NVIDIA CUDA Driver, Toolkit, cuDNN, and TensorRT 
04. Install the Jupyter Notebook Server
05. Install Virtual Environments in Jupyter Notebook
06. Install the Python Environment for AI and Machine Learning
07. Install Ubuntu Desktop With a Graphical User Interface (Bonus)
Windows 10:
01. Install and Manage Multiple Python Versions
02. Install the NVIDIA CUDA Driver, Toolkit, cuDNN, and TensorRT
03. Install the Jupyter Notebook Server
04. Install Virtual Environments in Jupyter Notebook
05. Install the Python Environment for AI and Machine Learning
MacOS:
01. Install and Manage Multiple Python Versions
02. Install the Jupyter Notebook Server
03. Install Virtual Environments in Jupyter Notebook
04. Install the Python Environment for AI and Machine Learning