NCP-AIO Troubleshooting and Optimization Practice Question
A data scientist reports that a Jupyter notebook running on a GPU-enabled server is extremely slow when training a small neural network, even though nvidia-smi shows the GPU is idle. The notebook uses TensorFlow. Which is the most likely cause?
⚠ Common exam trap
The trap here is assuming the GPU is too busy or the model too small, when the real issue is that TensorFlow is not configured to use the GPU at all.
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
✓
TensorFlow is not configured to use the GPU; it is running on the CPU.
The idle GPU during training strongly suggests TensorFlow is not using it. Common causes include missing GPU support in TensorFlow, incorrect CUDA/cuDNN versions, or the process not having access to the GPU. Verifying TensorFlow's device configuration is the first step. Other options are less likely given the evidence.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
TensorFlow is not configured to use the GPU; it is running on the CPU.
Why this is correct
If the GPU is idle while training, TensorFlow is likely defaulting to CPU execution. This can happen if the GPU is not visible to TensorFlow due to missing CUDA libraries, incorrect environment variables, or a CPU-only TensorFlow installation. Checking tf.config.list_physical_devices('GPU') would confirm. This is a common oversight in notebook environments where the kernel may not have GPU access.
- ✗
The GPU is being used by another process, causing contention.
Why it's wrong here
If another process were using the GPU, nvidia-smi would show utilization and memory usage. The scenario states the GPU is idle, so contention is not the issue. This option misinterprets the evidence. Contention would typically show high GPU utilization, not idle, so it is not the likely cause here.
- ✗
The Jupyter notebook kernel needs to be restarted to detect the GPU.
Why it's wrong here
While restarting the kernel might help if the environment was changed, it is not the most likely cause. The issue is more fundamental: TensorFlow may not be set up to use the GPU. Restarting without addressing configuration would not help. This option suggests a transient issue rather than a configuration problem.
- ✗
The neural network is too small to benefit from GPU acceleration, so TensorFlow automatically uses the CPU.
Why it's wrong here
TensorFlow does not automatically switch to CPU based on model size; it uses the GPU if available and configured. Even small models can benefit from GPU acceleration, though the speedup may be less pronounced. The idle GPU indicates no GPU usage at all, not a deliberate choice based on size. This explanation is technically incorrect.
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.