NCP-AIO Troubleshooting and Optimization Practice Question
A data scientist reports that a Jupyter notebook running on a DGX station cannot allocate GPU memory, even though other users' jobs are running fine. The notebook kernel was started before a system administrator updated the NVIDIA driver and rebooted the node. Which action should the data scientist take to resolve the issue?
⚠ Common exam trap
The trap here is assuming that a GPU reset or environment variable change is needed, when simply restarting the long-running process that predates the driver update is sufficient.
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
✓
Restart the Jupyter kernel to reinitialize CUDA and pick up the new driver.
After a driver update and reboot, any process that started before the update holds an outdated CUDA context. The Jupyter kernel is such a process, so it cannot use the new driver. Restarting the kernel creates a fresh process that loads the updated driver and can allocate GPU memory normally, resolving the issue without affecting other users.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Run `nvidia-smi --gpu-reset` to reset the GPU and clear any stale contexts.
Why it's wrong here
`nvidia-smi --gpu-reset` is used to reset a GPU that is in an error state, but it requires administrative privileges and can disrupt other running jobs. The issue here is a stale process context, not a GPU hardware error. Resetting the GPU is unnecessary and could cause outages for other users on the shared DGX station.
- ✗
Set the environment variable `CUDA_VISIBLE_DEVICES=0` to force the notebook to use a specific GPU.
Why it's wrong here
Setting `CUDA_VISIBLE_DEVICES` restricts which GPUs a process can see, but it does not resolve a driver version mismatch. The kernel would still be using the old driver context and would fail to allocate memory. This variable is useful for GPU isolation, not for fixing driver update issues in an already-running process.
- ✓
Restart the Jupyter kernel to reinitialize CUDA and pick up the new driver.
Why this is correct
When the NVIDIA driver is updated and the system reboots, any existing processes that had already initialized CUDA hold references to the old driver. The Jupyter kernel is such a process. Restarting the kernel terminates the old process and starts a new one that loads the updated driver, allowing CUDA calls to succeed and GPU memory to be allocated.
- ✗
Reinstall the NVIDIA driver using the `.run` installer without rebooting.
Why it's wrong here
Reinstalling the driver without rebooting would not help because the running kernel and processes still use the old driver modules. The system administrator already updated the driver and rebooted, so the correct driver is active. Reinstalling could introduce inconsistencies and is not a user-level fix for a stale kernel process.
Visual reference
About these practice questions
One of 309 original NCP-AIO practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. 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.