Greetings to all,
I have created two guides on how to bootstrap your own Nvidia Jetson-based AI and HPC clusters. These guides should provide a solid foundation for those interested in exploring and experimenting with SLURM and Kubernetes. Whether you are a hobbyist, researcher, or ML specialist looking to leverage the power of AI on an edge cluster, these guides should be helpful.
- “Running a ChatGPT-Like LLM-LLaMA2 on a Nvidia Jetson Cluster using Kubernetes” (Running a ChatGPT-Like LLM-LLaMA2 on a Nvidia Jetson Cluster - Hackster.io)
- “How to Build an Nvidia Jetson HPC Cluster Using SLURM” (How to build Nvidia Jetson HPC Cluster Using SLURM - Hackster.io)
Below the image of the last version of my cluster. The compute nodes are based on the Nvidia Jetson boards(Nvidia Jetson Nano 4GB, two Nvidia Jetson Xavier NX with 8GB and 16GB memory. A Raspberry Pi 4B acts as the head node, and an Orange Pi 5 Plus with 128GB SSD is used as the NFS node.
I hope these guides come in handy as you build your home cluster.