# Tutorial Install GPU and Deep Learning Frameworks
# Upgrade Kernel Ubuntu

sudo add-apt-repository ppa:cappelikan/ppa
sudo apt-get update && sudo apt-get upgrade
sudo apt install mainline

Open mainline app and install the last release kernel

Or compile your kernel

sudo apt-get install git fakeroot build-essential ncurses-dev xz-utils libssl-dev bc flex libelf-dev bison
latest_tag=$(git ls-remote --tags --sort='v:refname' https://github.com/torvalds/linux.git | tail -n1 | sed 's/.*\///; s/\^{}//')
kernel_version=$(echo $latest_tag | sed 's/^v//')

git clone --depth 1 --branch $latest_tag https://github.com/torvalds/linux.git "linux-$kernel_version"
cd "linux-$kernel_version"

# Crear la carpeta de salida si no existe
mkdir -p ./output

cp -v /boot/config-$(uname -r) .config
yes "" | scripts/config --disable SYSTEM_TRUSTED_KEYS
yes "" | scripts/config --disable SYSTEM_REVOCATION_KEYS
yes "" | make oldconfig

sudo make -j $(nproc)
sudo make -j $(nproc) modules_install
sudo make -j $(nproc) install 
# make -j $(nproc) deb-pkg

# Mover los archivos generados a la carpeta de salida
#mv ../*.deb ./output/
# cd ./output/

# sudo dpkg -i ../linux-headers-*.deb ../linux-image-*.deb

sudo update-initramfs -c -k $kernel_version  
sudo update-grub

# 1. Install Graphic Cards Drivers

## 1.1 Not necessary in MacOS

## 1.2 Case Nvidia:
### Download drivers from [Nvidia page]

sudo apt-get remove --purge '^nvidia-.*' --yes
sudo apt-get --purge remove nvidia* --yes
sudo apt-get --purge remove "*cublas*" "cuda*" --yes
sudo apt remove nvidia-* --yes
sudo apt-get --purge remove "*nvidia*" --yes
sudo apt autoremove --yes
sudo apt-get update --yes && sudo apt-get full-upgrade --yes
wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2404/x86_64/cuda-keyring_1.1-1_all.deb
sudo dpkg -i cuda-keyring_1.1-1_all.deb
sudo apt-get update
sudo apt-get install -y nvidia-open
sudo apt-get install nvitop --yes

curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg \
  && curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list | \
    sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \
    sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list \
  && \
    sudo apt-get update
sudo apt-get install -y nvidia-container-toolkit

### In case of GTX

sudo apt-get install -y cuda-drivers
echo "options nvidia NVreg_OpenRmEnableUnsupportedGpus=1" | sudo tee /etc/modprobe.d/nvidia-gsp.conf

## 1.3 Case Intel

# Intel APT
sudo apt update
sudo apt -y install cmake pkg-config build-essential

sudo wget -O- https://apt.repos.intel.com/intel-gpg-keys/GPG-PUB-KEY-INTEL-SW-PRODUCTS.PUB \ | gpg --dearmor | sudo tee /usr/share/keyrings/oneapi-archive-keyring.gpg > /dev/null && echo "deb [signed-by=/usr/share/keyrings/oneapi-archive-keyring.gpg] https://apt.repos.intel.com/oneapi all main" | sudo tee /etc/apt/sources.list.d/oneAPI.list

# CODEPLAY
sudo wget -qO - https://developer.codeplay.com/apt/public.key | gpg --dearmor | sudo tee /usr/share/keyrings/codeplay-keyring.gpg > /dev/null && echo "deb [signed-by=/usr/share/keyrings/codeplay-keyring.gpg] https://developer.codeplay.com/apt all main" | sudo tee /etc/apt/sources.list.d/codeplay.list

### Restart the PC

## 1.4 Case AMD

sudo apt-get remove --purge '^rocm-.*'
sudo apt remove rocm-*
sudo apt purge '*rocm*'
sudo apt update && sudo apt full-upgrade
sudo apt install "linux-headers-$(uname -r)" "linux-modules-extra-$(uname -r)"

wget https://repo.radeon.com/amdgpu-install/7.0.1/ubuntu/noble/amdgpu-install_7.0.1.70001-1_all.deb

sudo apt install ./amdgpu-install_7.0.1.70001-1_all.deb
sudo apt update
sudo apt install python3-setuptools python3-wheel
sudo apt install amdgpu-dkms # drivers
sudo apt install rocm

sudo usermod -a -G render,video $LOGNAME # Add the current user to the render and video groups

### Restart the PC

# 2. Install Library HPC

## 2.1 Case Nvidia

### 2.1.2 Install Cuda:
Source: https://developer.nvidia.com/cuda-downloads

Extra information:
The **cuda** and **cuda-toolkit** meta-packages will be equivalent in CUDA 12.8.

sudo apt-get -y install cuda-toolkit cudnn tensorrt nvidia-gds

Add PATH to .bashrc

sudo gedit ~/.bashrc

export PATH=/usr/local/cuda-13.0/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda-13.0/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

source ~/.bashrc

### Restart the PC

## 2.2 Case Intel

## Install oneAPI:

# Install all the packages or selected packages.
All basic compilers (Fortran, C, C++, and Python)
sudo apt install intel-basekit

# HPC toolkits
sudo apt install intel-hpckit

# IoT (internet of things) toolkits
sudo apt install intel-iotkit

# AI (Artificial Intelligence) analytics toolkits
sudo apt install intel-aikit

# Rendering toolkit
sudo apt install intel-renderkit

# After Installation
# After you have installed the oneAPI kits, they can be activated using the following command written in the ~/.bashrc file.

# For all users (root access)
. /opt/intel/oneapi/setvars.sh

# For single user
. ~/intel/oneapi/setvars.sh

### Restart the PC

## 2.3 Case AMD
### Install RocM:

Add PATH to .bashrc

sudo gedit ~/.bashrc

export ROCM_PATH=/opt/rocm-6.4.3
export PATH=$ROCM_PATH/bin:$PATH
export LD_LIBRARY_PATH=$ROCM_PATH/lib:$LD_LIBRARY_PATH

source ~/.bashrc

# check with rocminfo
rocminfo

amd-smi metric

## 3. Install Anaconda

## Case Linux
1. Download .sh: https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
2. ```chmod +x Miniconda3-latest-Linux-x86_64.sh```
3. ```bash Miniconda3-latest-Linux-x86_64.sh```

## Case Mac
1. Download .sh: https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-arm64.sh
2. ```chmod +x Miniconda3-latest-MacOSX-arm64.sh```
3. ```bash Miniconda3-latest-MacOSX-arm64.sh```

# Benchmark
1. Create your environment ```conda create -n myenv python=3.11 & conda activate myenv```
2. Install framework tensorflow or pytorch
3. Install jupyter lab in console: ```conda install numpy pandas matplotlib notebook jupyterlab```
4. Open jupyter lab in console: ```jupyter lab```
5. Create a notebook
6. Install ai-benchmark in a cell: ```!pip3 install new-ai-benchmark```
7. Execute:

from ai_benchmark import AIBenchmark
benchmark = AIBenchmark(use_cpu=False, verbose_level=2) # FOR CPU use_cpu=True
results = benchmark.run()

## Optional Install Tensorflow...

### Nvidia

pip install tensorflow[and-cuda]
# Verify install:
python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))"

### Extension Intel

pip install --upgrade intel-extension-for-tensorflow[gpu]

### Apple

python -m pip install tensorflow
python -m pip install tensorflow-metal

### AMD

pip install -U tensorflow-rocm

# Optional: Install Pytorch

### Case Nvidia

### Case Intel

pip3 install -U --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cpu
python -m pip install intel_extension_for_pytorch

### Case Mac

### Case AMD

If you are using cuda <12.0

import tensorflow as tf
gpus = tf.config.experimental.list_physical_devices('GPU')
for gpu in gpus:
tf.config.experimental.set_memory_growth(gpu, True)

-----------------------------------------------------------

## FIX NUMA ERROR

sudo echo 0 | sudo tee -a /sys/bus/pci/devices/0000\:01\:00.0/numa_node