This repo demonstrates the setup of all required Intel GPU and NPU user-space drivers for use with Intel Extensionf for Pytorch. It has been tested on a Intel 13th Gen Intel(R) Core(TM) i9-13900H, Intel(R) Core(TM) Ultra 5 125H processor, and 13th Gen Intel(R) Core(TM) i7-13700K with Intel(R) ARC A770 Discrete GPU
- Ubuntu 24.04 with kernel 6.8.0-36-generic (for Intel GPU kernel driver support)
- Installed Docker (for Linux)
- Intel ARC series GPU (tested with Intel ARC A770 16GB and Intel(R) Core(TM) Ultra 5 125H integrated GPU)
These samples utilize containers to fully encapsulate the example with minimial host dependencies. Here are the instructions how to install docker:
$ sudo apt-get update
$ sudo apt-get install ca-certificates curl
$ sudo install -m 0755 -d /etc/apt/keyrings
$ sudo curl -fsSL https://download.docker.com/linux/ubuntu/gpg -o /etc/apt/keyrings/docker.asc
$ sudo chmod a+r /etc/apt/keyrings/docker.asc
# Add the repository to Apt sources:
echo \
"deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/docker.asc] https://download.docker.com/linux/ubuntu \
$(. /etc/os-release && echo "$VERSION_CODENAME") stable" | \
sudo tee /etc/apt/sources.list.d/docker.list > /dev/null
sudo apt-get update
# Install docker
sudo apt-get install docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin
# Enable docker access as user
sudo groupadd docker
sudo usermod -aG docker $USER
To build the container, type:
$ ./build
To execute a sample app to print information about GPU devices:
$ ./run
- Kernel 6.8.0-38-generic and later may cause a ARC GPU hang. See https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2072755. Use Kernel 6.8.0-36-generic or older