NCP-AIO Workload Management Practice Question
A Kubernetes cluster running the NVIDIA GPU Operator is shared by an inference team and a research team. The research team's training pods repeatedly evict the inference pods from GPUs, causing latency spikes in production. The administrator wants to guarantee that inference pods always get GPU access first. Which Kubernetes scheduling mechanism should be configured?
⚠ Common exam trap
The trap here is assuming that taints, tolerations, or PodDisruptionBudgets control scheduler contention, when in fact only PriorityClass with preemption orders competing pods for scarce GPU resources.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Assign a higher PriorityClass to the inference pods and enable preemption on the scheduler.
Priority and preemption are the native Kubernetes controls that decide which pods win when GPU resources are contended. Giving inference pods a higher PriorityClass lets the scheduler admit them first and evict lower-priority training pods when necessary, which directly satisfies the requirement that production inference always obtains GPU access.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Enable the NVIDIA MIG feature on all GPUs and dedicate a MIG instance to each inference pod.
Why it's wrong here
MIG partitions a physical GPU into isolated slices, which improves isolation and predictability but does not control which pods win scheduling contention. Without a priority mechanism, training pods could still occupy the MIG slices needed by inference pods, so MIG alone does not guarantee inference-first access.
- ✓
Assign a higher PriorityClass to the inference pods and enable preemption on the scheduler.
Why this is correct
PriorityClass with preemption allows higher-priority inference pods to be scheduled and to evict lower-priority training pods when GPU resources are scarce. This directly protects production inference latency by ensuring inference workloads are admitted first, which matches the requirement to guarantee GPU access for the inference team.
- ✗
Configure a taint on the GPU nodes and add the corresponding toleration only to the inference pods.
Why it's wrong here
A taint with a matching toleration prevents training pods from being scheduled on GPU nodes at all, which would block the research team entirely rather than prioritizing inference. The scenario requires both teams to share GPUs, so excluding one team is not the intended outcome and does not address eviction ordering.
- ✗
Create a PodDisruptionBudget for the inference deployment and set maxUnavailable to zero.
Why it's wrong here
A PodDisruptionBudget limits voluntary disruptions such as node drains, not scheduler-driven preemption or eviction caused by higher-priority workloads. It would not stop training pods from consuming GPU resources or from being scheduled ahead of inference pods, so it fails to guarantee inference pod priority.
About these practice questions
This NCP-AIO question is part of Courseiva's 309-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official NVIDIA exam blueprint
This NCP-AIO practice question is part of Courseiva's free NVIDIA certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the NCP-AIO exam.