NVIDIA Software Stack & Developer Tools Flashcards

What is the primary purpose of the cuDNN library in the NVIDIA software stack?

Answer: Providing optimized primitives for deep neural network operations

cuDNN (CUDA Deep Neural Network library) provides highly tuned implementations of primitives such as convolution, pooling, normalization, and activation functions used in deep learning.

Which NVIDIA library provides GPU-accelerated implementations of the Basic Linear Algebra Subprograms (BLAS) for dense matrices?

Answer: cuBLAS

cuBLAS is the NVIDIA CUDA implementation of BLAS, providing optimized routines for dense vector and matrix operations on NVIDIA GPUs.

What does NCCL stand for in the NVIDIA ecosystem?

Answer: NVIDIA Collective Communications Library

NCCL (NVIDIA Collective Communications Library) implements multi-GPU and multi-node collective operations such as AllReduce, Broadcast, and AllGather, optimized for NVIDIA interconnects.

What best describes the NVIDIA RAPIDS suite?

Answer: A GPU-accelerated open-source data science and analytics library suite

RAPIDS is an open-source suite of GPU-accelerated libraries (cuDF, cuML, cuGraph, etc.) that enables end-to-end data science and analytics workflows on NVIDIA GPUs.

Which NVIDIA library is specifically designed to accelerate data loading and augmentation pipelines for deep learning training?

Answer: DALI (Data Loading Library)

NVIDIA DALI (Data Loading Library) is a portable, high-performance library that decodes images, videos, and audio, and performs data augmentation directly on the GPU to eliminate CPU bottlenecks.

What is the primary function of the nvidia-smi command-line utility?

Answer: Monitoring and managing NVIDIA GPU devices and their status

nvidia-smi (System Management Interface) provides real-time monitoring of GPU utilization, memory usage, temperature, power draw, and running processes, and supports management actions like setting power limits.

How does cuSPARSE differ from cuBLAS in the NVIDIA software stack?

Answer: cuSPARSE provides optimized routines for sparse matrix formats; cuBLAS handles dense matrices

cuSPARSE provides GPU-accelerated operations for sparse matrices stored in compressed formats (CSR, CSC, COO), while cuBLAS targets dense matrix and vector operations.