NCP-AIO Workload Management Practice Question
When running multi-instance GPU (MIG) workloads, what is the main advantage of assigning specific MIG profiles to different Kubernetes namespaces?
⚠ Common exam trap
Candidates often assume MIG is primarily for performance tuning or cost optimization, missing that its fundamental architectural strength in Kubernetes is hardware-level memory isolation between distinct, potentially insecure, user workloads.
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
✓
It enhances GPU memory security by isolating workload address spaces.
MIG profiles allow for fine-grained hardware isolation, ensuring that different workloads—such as small inference tasks and large training runs—do not interfere with each other. By mapping profiles to specific namespaces, administrators can enforce strict hardware separation, preventing 'noisy neighbor' issues where one workload consumes memory bandwidth or compute cycles required by another, thus improving overall cluster stability and predictability.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
It allows the GPU to switch between different CUDA versions per instance.
Why it's wrong here
MIG provides hardware-level isolation of compute and memory, but the underlying CUDA driver version remains the same across all instances on the physical GPU. It does not provide the ability to run multiple, conflicting driver versions simultaneously on a single GPU, as the hardware requires a single driver instance.
- ✓
It enhances GPU memory security by isolating workload address spaces.
Why this is correct
MIG profiles provide hardware-level isolation of memory and compute resources. By assigning specific profiles to namespaces, you effectively prevent cross-talk between different workloads' address spaces. This is a critical security and operational feature for multi-tenant environments, ensuring that sensitive data in one inference pod cannot be accessed by another.
- ✗
It automatically compresses the model weights for faster loading.
Why it's wrong here
MIG is a hardware-partitioning technology, not a model optimization tool. It does not perform weight compression, quantization, or any other form of model-level processing. Its function is limited to hardware resource allocation and does not influence the data representation or storage format of the neural networks it supports.
- ✗
It allows the cluster to bypass standard Kubernetes scheduler policies.
Why it's wrong here
MIG works in conjunction with the Kubernetes scheduler. It does not bypass scheduler policies; instead, it provides the scheduler with more granular resource requests (e.g., requesting a specific MIG profile). Bypassing policies would lead to unmanageable clusters and is not a supported or intended use case for MIG.
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.