Setting up TensorRT Environment on Ubuntu 22.04 / 20.04

Summary

Ubuntu 22.04 or Ubuntu 20.04

NVIDIA driver version: 535 (535.129.03)

CUDA version: 12.1.1

cuDNN version: 8.9.7 for CUDA 12.X

TensorRT version: 8.6 GA

Installing CUDA

Follow the official instruction, download CUDA from here.

Ubuntu 22.04

Select CUDA 12.4.0 (March 2024), and then execute the commands prompted by the instruction on the website.

Here's an example set of commands to run:

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-ubuntu2204.pin
sudo mv cuda-ubuntu2204.pin /etc/apt/preferences.d/cuda-repository-pin-600
wget https://developer.download.nvidia.com/compute/cuda/12.4.0/local_installers/cuda-repo-ubuntu2204-12-4-local_12.4.0-550.54.14-1_amd64.deb
sudo dpkg -i cuda-repo-ubuntu2204-12-4-local_12.4.0-550.54.14-1_amd64.deb
sudo cp /var/cuda-repo-ubuntu2204-12-4-local/cuda-*-keyring.gpg /usr/share/keyrings/
sudo apt-get update
sudo apt-get -y install cuda-toolkit-12-4

Ubuntu 20.04

After installation, add CUDA to ~/.bashrc:

# ~/.bashrc
...

# CUDA
export PATH="/usr/local/cuda-12.4/bin/:$PATH"
export LD_LIBRARY_PATH="/usr/local/cuda-12.4/lib64/:$LD_LIBRARY_PATH"

...

To test cuda installation, run the following command. The system should be able to find nvcc.

$ nvcc -V
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2024 NVIDIA Corporation
Built on Tue_Feb_27_16:19:38_PST_2024
Cuda compilation tools, release 12.4, V12.4.99
Build cuda_12.4.r12.4/compiler.33961263_0

Note:

nvidia-smi might fail to run with the following error:

$ nvidia-smi
Failed to initialize NVML: Driver/library version mismatch

This happens when the CUDA we installed is a different version than the one comes with the driver. If this happens, reboot the system to let nvidia-smi reload the correct CUDA.

After reboot:

tk@DESKTOP-Scratch:~$ nvidia-smi
Fri Dec 22 13:48:54 2023       
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 530.30.02              Driver Version: 530.30.02    CUDA Version: 12.1     |
|-----------------------------------------+----------------------+----------------------+
| GPU  Name                  Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf            Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                                         |                      |               MIG M. |
|=========================================+======================+======================|
|   0  NVIDIA GeForce GTX 1070         On | 00000000:27:00.0 Off |                  N/A |
| 34%   37C    P8                8W / 190W|    178MiB /  8192MiB |      1%      Default |
|                                         |                      |                  N/A |
+-----------------------------------------+----------------------+----------------------+
                                                                                         
+---------------------------------------------------------------------------------------+
| Processes:                                                                            |
|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
|        ID   ID                                                             Usage      |
|=======================================================================================|
|    0   N/A  N/A       987      G   /usr/lib/xorg/Xorg                           44MiB |
|    0   N/A  N/A      1259      G   /usr/bin/gnome-shell                        131MiB |
+---------------------------------------------------------------------------------------+

Installing cuDNN

Follow the official instruction, download cuDNN here.

Ubuntu 22.04

Select "cuDNN v9.3.0, for CUDA 12.x" with the .deb file option, and then execute the commands prompted by the instruction on the website.

Here's an example set of command to run:

wget https://developer.download.nvidia.com/compute/cudnn/9.3.0/local_installers/cudnn-local-repo-ubuntu2204-9.3.0_1.0-1_amd64.deb
sudo dpkg -i cudnn-local-repo-ubuntu2204-9.3.0_1.0-1_amd64.deb
sudo cp /var/cudnn-local-repo-ubuntu2204-9.3.0/cudnn-*-keyring.gpg /usr/share/keyrings/
sudo apt-get update
sudo apt-get -y install cudnn

Ubuntu 20.04

sudo apt install -y cudnn-cuda-12

(might not need do the following)

Install the runtime library.

sudo apt-get install libcudnn8=8.9.7.29-1+cuda12.2

Install the developer library.

sudo apt-get install libcudnn8-dev=8.9.7.29-1+cuda12.2

Install the code samples.

sudo apt-get install libcudnn8-samples=8.9.7.29-1+cuda12.2

run the test program to see if success

https://forums.developer.nvidia.com/t/verify-cudnn-install-failed/167220/4

Installing TensorRT

There are two parts of TensorRT installation

TensorRT GA

Goto https://developer.nvidia.com/tensorrt

Download both the "TensorRT 10.3 GA for Linux x86_64 and CUDA 12.0 to 12.5 TAR Package" and the DEB package

Warning: Be careful to download the version matching with system CUDA version.

Install the DEB package with Software Install.

Alternatively, do the following commands

sudo dpkg -i ./nv-tensorrt-local-repo-ubuntu2204-10.3.0-cuda-12.5_1.0-1_amd64.deb
sudo cp /var/nv-tensorrt-local-repo-ubuntu2204-10.3.0-cuda-12.5/nv-tensorrt-local-620E7D29-keyring.gpg /usr/share/keyrings/

sudo apt update
sudo apt install nv-tensorrt-local-repo-ubuntu2204-10.3.0-cuda-12.5

We also need to link the libraries. Unpack the tar package:

tar xzvf ./TensorRT-10.3.0.26.Linux.x86_64-gnu.cuda-12.5.tar.gz

Then. move the unpacked directory to the installation path (~/Documents/), and add to bashrc

...

# TensorRT
export TRT_LIBPATH="/home/tk/Documents/TensorRT-10.3.0.26/targets/x86_64-linux-gnu/lib/"
export LD_LIBRARY_PATH="/home/tk/Documents/TensorRT-10.3.0.26/lib/:$TRT_LIBPATH:$LD_LIBRARY_PATH"

...

Install to Python using the following command

cd ~/Documents/TensorRT-10.3.0.26/python/
pip install ./tensorrt-10.3.0-cp38-none-linux_x86_64.whl

For TensorRT < 10.3, it might also need this dependency

cd ~/Documents/TensorRT-8.6.1.6/graphsurgeon/
pip install ./graphsurgeon-0.4.6-py2.py3-none-any.whl

TensorRT OSS

Seems that normally we don't need this thing.

For python:

git clone -b main https://github.com/nvidia/TensorRT TensorRT
cd TensorRT
git submodule update --init --recursive

 cd TensorRT
 mkdir -p build && cd build

cmake .. -DTRT_LIB_DIR=$TRT_LIBPATH/lib/ -DTRT_OUT_DIR=`pwd`/out

make -j$(nproc)

Other Dependencies

pip3 install cuda-python

sudo apt install libssl-dev

sudo apt install cmake g++

$ cmake --version
cmake version 3.16.3

CMake suite maintained and supported by Kitware (kitware.com/cmake).

pip install onnx
pip install onnxruntime
pip install onnxruntime-gpu
pip install onnx-simplifier

Final Step

Follow this repo to use TRT

trt-samples-for-hackathon-cn/cookbook at main · T-K-233/trt-samples-for-hackathon-cnGitHub

FAQ

Hit:7 http://us.archive.ubuntu.com/ubuntu focal-updates InRelease
Err:3 file:/var/nv-tensorrt-local-repo-ubuntu2004-8.6.1-cuda-12.0  InRelease
  The following signatures couldn't be verified because the public key is not available: NO_PUBKEY 6694DE8A9A1EDFBA
Hit:8 http://us.archive.ubuntu.com/ubuntu focal-backports InRelease
Reading package lists... Done
W: GPG error: file:/var/nv-tensorrt-local-repo-ubuntu2004-8.6.1-cuda-12.0  InRelease: The following signatures couldn't be verified because the public key is not available: NO_PUBKEY 6694DE8A9A1EDFBA
E: The repository 'file:/var/nv-tensorrt-local-repo-ubuntu2004-8.6.1-cuda-12.0  InRelease' is not signed.
N: Updating from such a repository can't be done securely, and is therefore disabled by default.
N: See apt-secure(8) manpage for repository creation and user configuration details.

solution:

the key is not installed. run the command in "Installing TensorRT" again

sudo cp /var/nv-tensorrt-local-repo-ubuntu2204-8.6.1-cuda-12.0/nv-tensorrt-local-42B2FC56-keyring.gpg /usr/share/keyrings/