Join our Discord Server
📗
New Book: Docker AI
– Learn to build & deploy AI-powered apps with Docker!
Get the Book →
×
Follow
Collabnix
Home
Learn
Resources
Docker
Kubectl
Docker Model Runner
Terraform
ChatGPT Cheat Sheet
Ansible Cheatsheet
Helm Cheatsheet [ 2024 Updated]
OpenShift CheatSheet
Claude Code Cheatsheet
Ollama Cheatsheet [2026 Updated]
KubeLabs
DockerLabs
Terraform Labs
Raspberry Pi
Jetson Nano
Jetson AGX Xavier
Cheatsheets
Community
Events
Chat
Slack
Discord
Write for Us!
DGX Spark
NVIDIA DGX Spark + Kubernetes: Run GPU Workloads on the GB10 Grace Blackwell Superchip
The NVIDIA DGX Spark puts a 1 PFLOP (FP4) Grace Blackwell GB10 superchip and 128 GB of unified memory on your...
Join our Discord Server
Table of Contents
×
Hardware & software baseline
Step 0 — Confirm the GPU works on the host
Step 1 — Point containerd at the NVIDIA runtime
Step 2 — Install K3s
Step 3 — The unified-memory gotcha (and the fix)
Step 4 — Install the GPU Operator (driver + toolkit pre-installed)
Step 5 — Smoke test schedule a GPU pod
Step 6 — Share one GPU across many pods (time-slicing)
Step 7 — Deploy a real workload vLLM serving Qwen3
Step 8 — GPU observability with DCGM exporter
Troubleshooting cheat sheet
FAQ
Wrapping up
Related Posts
→
Index
