Skip to content
AI Security

Interview question · Cloud, Container & AI Infrastructure Security

How do you run vulnerability and patch management for a GPU fleet used for training and inference?

GPU nodes carry layers most patch programmes forget: drivers, firmware, CUDA libraries and container toolkits. The scored answer inventories them, prioritises by exposure, and patches without breaking month-long training runs.

mediumNewPreparation tags: NVIDIA, Google, AmazonUpdated Nov 2026

GPU nodes carry layers most patch programmes forget: drivers, firmware, CUDA libraries and container toolkits. The scored answer inventories them, prioritises by exposure, and patches without breaking month-long training runs.

Make your first attempt

What would you say before reading the answer?

  1. 01 State the assumptions your answer depends on.
  2. 02 Explain the mechanism and choose a control.
  3. 03 Name a condition that would change your decision.

The worked answer includes an approach, scoring criteria and follow-up questions. Compare the reasoning with your own.

10 answers per topic instead of 5, plus saved progress and bookmarks · costs nothingor unlock all 529 remaining answers · $25

Every worked answer in all 17 topics, every lesson in every course, the full concept curriculum, the complete practice-test bank, and the downloadable security field guides.

6 months · One payment · No auto-renewal

What if · the follow-ups in this answer

The answer works through 6 follow-ups like an interviewer would

A critical GPU driver vulnerability is announced, and the company runs hundreds of GPU nodes for training and multi-tenant inference.

  1. What must be inventoried on GPU nodes?
  2. Where do advisories come from?
  3. Which nodes go first?
  4. How do you patch without breaking training runs?
  5. What if patching takes weeks?
  6. How would you run the patch?

Each rung has a nudge, a model answer, the principle that carries over and the weak answer to avoid. Premium opens every answer and ladder.

UP NEXT ON YOUR JOURNEY
Next in this trackSecure the container image pipeline for model serving.Next in this trackDesign private networking and egress for an application that calls hosted model APIs.Next in this trackThreat-model the path a model takes from training checkpoint to serving node.
DISCUSSION · 0

No comments yet. Share a correction or a different way to answer it.