~/talk

Tutorial: Build, Operate, and Use a Multi-Tenant AI Cluster Based Entirely on Open Source

name:C. Misale
links:none on file
links:none on file
name:D. Grove
links:none on file
slides:none on file

Open-source stack shows a concrete recipe: deploy Nvidia GPU Operator, Kueue for workload-aware queuing and quotas, scheduler plugins for gang and topology-aware placement, and app wrapper plus autopilot for automated health checks and fault recovery. Demos run on a 24‑GPU test cluster and IBM Research reports operating ~1,200 GPUs with near-100% utilization by letting teams borrow unused quota, preempting or suspending jobs, and automatically migrating workloads off failing GPUs without human intervention.

AI-written summary · about

  1. 0:00Multi-tenant AI GPU clusters
  2. 2:25AI workloads and hardware needs
  3. 6:20Kubernetes and platform overview
  4. 10:32Cluster setup prerequisites
  5. 12:17Cluster config and team onboarding
  6. 18:18Queuing and quota with Kueue
  7. 20:03Kueue demo and preemption
  8. 32:14Fault detection with Autopilot
  9. 47:59Fault recovery with App Wrappers
  10. 55:01Kubeflow trainer example
  11. 57:32KubeRay fine-tuning example
  12. 1:06:28Batch inference example
  13. 1:09:59Key takeaways and results
~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.