Giters

rapidsai / node

GPU-accelerated data science and visualization in node

Home Page:https://rapidsai.github.io/node/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Package?

nicolasgere opened this issue · comments

commented

HI there, thanks for the amazing work!
We are very interested to use the webgl package, we are on headless-gl for now, but it does not work with gpu. Do you have an ETA on when we could start to test? Thank you!

commented

@nicolasgere you can test it right now by installing it via docker. You can find the published containers here. The package should listed as @nvidia/webgl, although #313 will update it to @rapidsai/webgl. Let us know if you face any issues!

commented

Do you mean that if I npm install @nvidia/webgl inside the container, it will install the package correctly in my node module?

commented

No actually, the package should come pre-installed in that docker image. You would just use it directly!

commented

where can I found a demo for webgl? I am facing the error 2.1.0 Missing GL version

commented

no update here @AjayThorve ?

commented

@nicolasgere are you running the WebGL demo from inside a docker container? If so, are you passing the --runtime=nvidia flag (and have the nvidia-container-toolkit installed)?

Our dev and runtime docker-compose files are configured to run GL apps inside the container but render in X11 on the host.

You can pass the same configuration via docker run. Save the following as a file called run-node-rapids-demo.sh:

Click here to see "run-node-rapids-demo.sh"
#!/usr/bin/env bash

envvars="\
-e NODE_NO_WARNINGS=1 \
`# Colorize the terminal in the container if possible` \
-e TERM=${TERM:-} \
`# Use the host's X11 display` \
-e DISPLAY=${DISPLAY:-} \
-e XAUTHORITY=${XAUTHORITY:-} \
-e NVIDIA_DRIVER_CAPABILITIES=all \
-e XDG_SESSION_TYPE=${XDG_SESSION_TYPE:-} \
-e XDG_RUNTIME_DIR=${XDG_RUNTIME_DIR:-/run/user/$UID} \
-e DBUS_SESSION_BUS_ADDRESS=${DBUS_SESSION_BUS_ADDRESS:-unix:path=/run/user/$UID/bus} \
"

volumes="\
-v /etc/fonts:/etc/fonts:ro \
-v /tmp/.X11-unix:/tmp/.X11-unix:rw \
-v /usr/share/fonts:/usr/share/fonts:ro \
-v /usr/share/icons:/usr/share/icons:ro \
-v /etc/timezone:/etc/timezone:ro \
-v /etc/localtime:/etc/localtime:ro \
-v /run/dbus/system_bus_socket:/run/dbus/system_bus_socket \
-v ${XDG_RUNTIME_DIR:-/run/user/$UID}:${XDG_RUNTIME_DIR:-/run/user/$UID} \
"

# Default to the "luma.gl-lessons/01" demo if none was passed in
cmd="${@:-npx @rapidsai/demo-luma.gl-lessons 01}"

exec docker run --rm -it --runtime=nvidia \
    ${envvars} \
    ${volumes} \
    ghcr.io/rapidsai/node:22.02.00-runtime-node16.13.2-cuda11.6.0-ubuntu20.04-demo \
    `# Run cmd inside the container` \
    ${cmd}
Running the above script should open a window on your host with a triangle and square drawn via OpenGL: image
Here's an example of running a different demo in the container with this script: image