NVIDIA-Certified-Professional Accelerated Data Science Sample Questions:

1. You are working on an MLOps pipeline with a large dataset using NVIDIA technologies. You need to assess the memory size of the dataset before training the model to avoid memory overflow errors.
Which of the following is the best method for verifying the memory size of the dataset on the GPU using NVIDIA tools?

A) Load the dataset into a CPU-based pandas dataframe and use the pandas.memory_usage() function to estimate the dataset's size before transferring it to the GPU.
B) Use NVIDIA DALI to process the dataset and directly query the memory size through DALI's DataLoader function.
C) Use cuDF to load the dataset into GPU memory and use the nvidia-smi tool to monitor GPU memory usage during dataset loading.
D) Use PyTorch's torch.cuda.memory_allocated() function to check the memory allocated on the GPU after the dataset is loaded.

2. You are deploying an NVIDIA GPU-accelerated machine learning model in a Docker container and want to ensure that your application can leverage the GPU efficiently.
What is the best way to manage CUDA dependencies and avoid compatibility issues inside your Docker container?

A) Use a base Ubuntu image and install TensorFlow, PyTorch, and CUDA using pip install inside the container.
B) Disable GPU acceleration in Docker and force computations on the CPU to avoid CUDA compatibility issues.
C) Manually install CUDA and cuDNN inside the container by downloading them from NVIDIA's website and setting environment variables.
D) Use NVIDIA's official Docker images from NVIDIA GPU Cloud (NGC), which come with pre-installed CUDA and AI frameworks.

3. You are tasked with acquiring a dataset for training a machine learning model in healthcare, predicting patient readmission rates. Before using the dataset, you must assess its quality.
Which of the following is the most important factor to evaluate before acquisition?

A) The number of missing values and inconsistencies in key columns
B) The programming language used to preprocess the dataset
C) The file format (CSV, JSON, or XML) of the dataset
D) Whether the dataset is stored on a cloud server or a local machine

4. A machine learning team needs to process terabytes of image metadata stored in a distributed storage system. They want to leverage GPU acceleration to speed up preprocessing and transformation while ensuring efficient parallel access.
Which of the following approaches best aligns with NVIDIA's accelerated data science ecosystem?

A) Use Dask-cuDF with Parquet files stored in an object storage system like S3.
B) Use SQLite for metadata storage and process it with RAPIDS cuDF.
C) Process metadata using PySpark on a CPU cluster before sending it to a GPU.
D) Store metadata in a traditional SQL database and query it using Pandas.

5. You are working on a large-scale graph analysis problem that involves computing the shortest paths between nodes in a massive social network dataset. You decide to leverage NVIDIA RAPIDS cuGraph for accelerated computation.
Which of the following cuGraph functions should you use?

A) cugraph.sssp()
B) cugraph.k_truss()
C) cugraph.pagerank()
D) cugraph.label_propagation()

Solutions: