Nest — NVIDIA cluster operations lab simulator

Unofficial practice simulator: BCM provisioning, Slurm scheduling, and Kubernetes GPU workloads in a real type-in terminal. 14 hands-on labs, a 51-question practice exam, live man pages, and per-task answer keys. Free, no signup.

Labs cover Slurm job submission and troubleshooting, Kubernetes GPU scheduling with the device plugin, BCM diagnostics (dcgmi, NVLink, InfiniBand), end-to-end provisioning, and node-down/role/Munge failure scenarios — on a stateful simulated DGX H100/H200 cluster.

Project Nest Egg overview · ROCnest — the AMD track

NEST
Unpacking...