NCA Containerization & NGC Catalog 3

Question 1

What is the function of NVIDIA's 'nvidia-container-runtime' in a containerized GPU environment?

Accepted Answer

It is an OCI-compliant runtime hook that injects GPU capabilities into containers

Answer

nvidia-container-runtime is an OCI-compliant container runtime that hooks into the standard container lifecycle to inject GPU driver libraries and devices into containers.

Question 2

When authenticating to the NGC container registry via Docker, which registry hostname is used?

Accepted Answer

nvcr.io

Answer

NGC container images are hosted at nvcr.io, so Docker login and pull commands must reference this hostname.

Question 3

Which environment variable is set inside NGC containers to indicate the minimum required CUDA driver version?

Accepted Answer

NVIDIA_REQUIRE_CUDA

Answer

The NVIDIA_REQUIRE_CUDA environment variable specifies the minimum CUDA and driver version required for the container to run correctly.

Question 4

In an NGC container, what does the environment variable CUDA_VISIBLE_DEVICES control?

Accepted Answer

Which physical GPUs are exposed to the container's processes

Answer

CUDA_VISIBLE_DEVICES restricts which host GPU devices are visible and accessible to processes running inside the container.

Question 5

What distinguishes an NGC 'AI Workflow' from a standard NGC container image?

Accepted Answer

AI Workflows bundle end-to-end reference pipelines including data preparation, training, and inference scripts

Answer

NGC AI Workflows are complete reference implementations that include all steps from data preprocessing through model training and deployment, not just a runtime environment.

(NCA) NVIDIA-Certified Associate Practice Test

(NCA) NVIDIA-Certified Associate Practice Test

NCA Containerization & NGC Catalog 3