NCA NVIDIA Software Stack & Developer Tools 2

Question 1

What distinguishes NVIDIA Nsight Compute from NVIDIA Nsight Systems?

Accepted Answer

Nsight Compute provides deep per-kernel GPU metrics; Nsight Systems delivers system-wide CPU+GPU timeline analysis

Answer

Nsight Compute profiles system-wide CPU+GPU timelines; Nsight Systems analyzes individual kernel metrics

Answer

Both tools perform identical analysis but target different GPU generations

Answer

Nsight Compute is for Windows only; Nsight Systems is cross-platform

Question 2

Which of the following is included in the NVIDIA CUDA Toolkit?

Accepted Answer

The nvcc compiler, cuBLAS, cuDNN, Nsight tools, and CUDA runtime libraries

Answer

Only the GPU driver and firmware

Answer

Pre-trained deep learning models and datasets

Answer

Only the nvcc compiler with no additional libraries

Question 3

What type of operations does the cuFFT library accelerate on NVIDIA GPUs?

Accepted Answer

Fast Fourier Transform computations

Answer

Federated Fine-Tuning of transformer models

Answer

File Format Translation between data types

Answer

Floating-point Fault Tolerance checks

Question 4

What is the role of nvcc in NVIDIA GPU development?

Accepted Answer

The NVIDIA CUDA compiler that compiles CUDA C/C++ source into GPU-executable binaries

Answer

A runtime daemon that schedules CUDA kernels on the GPU

Answer

A virtual machine for running CUDA code on CPUs

Answer

A container orchestration tool for NGC images

Question 5

Which capability makes NVIDIA Triton Inference Server particularly valuable for production AI deployments?

Accepted Answer

It serves models from multiple frameworks (TensorRT, ONNX, PyTorch, TF) under a single standardized API

Answer

It only supports TensorRT models for maximum performance

Answer

It trains new models from raw datasets in the cloud

Answer

It provides GUI-based model visualization and debugging

Question 6

What types of errors does the NVIDIA compute-sanitizer tool (formerly cuda-memcheck) detect?

Accepted Answer

GPU memory access violations, race conditions, and initialization errors in CUDA applications

Answer

Python syntax errors in CUDA wrapper code

Answer

Network packet loss during GPU-to-GPU data transfers

Answer

Power supply fluctuations causing GPU instability

Question 7

What is CUDA-GDB primarily used for in GPU application development?

Accepted Answer

Interactive debugging of CUDA applications at the thread and warp level

Answer

Automated GPU benchmark generation

Answer

Graphical display of GPU memory bandwidth utilization

Answer

Generating build dependency graphs for CUDA projects