Quiz połączony z przewodnikiem · Średni Poziom

Kontenery GPU z quizem dotyczącym zestawu narzędzi kontenerowych NVIDIA

Oddziel sterowniki GPU hosta od kontenerowych bibliotek CUDA i prześledź kroki w czasie wykonywania potrzebne do udostępnienia i sprawdzenia poprawności urządzeń NVIDIA.

Powiązane ścieżki prowadząceKontenery GPU Nvidia Zestaw narzędzi kontenerów
Pytanie 1 z 8

Which component normally manages the physical NVIDIA GPU at the kernel level?

Podgląd każdego pytania8 pytania

Odpowiedzi i wyjaśnienia pojawiają się w miarę rozwiązywania quizu.

  1. Which component normally manages the physical NVIDIA GPU at the kernel level?
    • The host driver
    • The CUDA container image alone
    • The model checkpoint
    • The Python package lockfile
  2. What role does NVIDIA Container Toolkit provide?
    • A replacement GPU kernel driver stored in each image
    • Runtime integration that exposes devices and driver interfaces to containers
    • A framework-neutral model accuracy benchmark
    • A compiler that converts all Python into CUDA
  3. Why can a container with CUDA libraries still fail to initialize the GPU?
    • CUDA libraries always replace the host driver
    • The container necessarily needs its own physical GPU
    • The host driver or runtime integration may be incompatible or missing
    • The model file does not contain a device node
  4. What does seeing a GPU in a container prove?
    • That memory allocation can never fail
    • That the model's accuracy is acceptable
    • That CUDA and driver versions are identical
    • Device visibility, but not that every application kernel will run correctly
  5. Why is updating CUDA packages inside an image different from updating the host driver?
    • Both update the same kernel module automatically
    • They are different software layers with different responsibilities
    • The container cannot contain any runtime libraries
    • Driver versions are determined by the model weights
  6. Which test is stronger than checking device visibility alone?
    • Inspect the model filename
    • Read the image tag
    • Run a small representative framework operation on the GPU
    • Confirm the container starts without a GPU request
  7. What should an incident record include for reproducibility?
    • Only the framework version and model architecture
    • Only the host driver and GPU model
    • Image tag and Python version without runtime configuration
    • Image digest, host driver, toolkit and runtime configuration
  8. What does GPU containerization not provide by itself?
    • A fully independent hardware driver environment
    • Packaged user-space dependencies
    • A way to request devices at launch
    • An image filesystem