GPU support for OpenFaaS Edge¶
At the time of speaking, OpenFaaS Pro on Kubernetes supports GPUs for functions via Profiles.
OpenFaaS Edge supports Nvidia GPUs for core services only.
Example:
services:
ollama:
image: docker.io/ollama/ollama:latest
command:
- "ollama"
- "serve"
volumes:
- type: bind
source: ./ollama
target: /root/.ollama
ports:
- "127.0.0.1:11434:11434"
gpus: all
deploy:
restart: always
Learn more: How to Protect Your Data with Self-Hosted LLMs and OpenFaaS Edge