Nsight Developer Tools

Type: Developer tools suite Tags: NVIDIA, Nsight, developer tools, CUDA, profiling, debugging, correctness, graphics, cloud, IDE, SDK Related: NVIDIA-CUDA, Nsight-Systems, Nsight-Compute, Nsight-Copilot, Nsight-Graphics, Nsight-Cloud, Nsight-Visual-Studio-Edition, Nsight-Visual-Studio-Code-Edition, Nsight-Integration, Nsight-JupyterLab-Extension, Nsight-Python, Nsight-Deep-Learning-Designer, Nsight-Aftermath-SDK, Nsight-Perf-SDK, Nsight-Eclipse-Plugins, CUDA-GDB, Compute-Sanitizer, CUPTI, NVTX, CUDA-Debugger-API, NVIDIA-Developer-Program Sources: https://docs.nvidia.com/nsight-developer-tools/index.html, https://developer.nvidia.com/tools-overview, https://developer.nvidia.com/tools-tutorials Last Updated: 2026-04-29

Summary

NVIDIA Nsight Developer Tools is the suite-level umbrella for NVIDIA tools used to build, debug, profile, and analyze GPU-accelerated software across desktop, mobile, embedded, cloud, cluster, and HPC environments. It is a hub page; use the linked tool pages for implementation details and current requirements.

Detail

NVIDIA’s current docs describe Nsight Developer Tools as a comprehensive tool suite spanning desktop and mobile targets. The developer tools overview organizes the suite around code development, debugging/correctness, profiling, tracing, APIs, platform analysis, kernel analysis, graphics analysis, cloud profiling, and SDK integration.

For profiling, start with Nsight-Systems for system-wide timeline analysis and then use Nsight-Compute for CUDA kernel-level metrics or Nsight-Graphics for graphics frame, shader, and GPU Trace workflows. Nsight-Copilot adds CUDA-aware AI assistance around VS Code development and preview Nsight Compute kernel-profiling guidance. For cloud and Kubernetes environments, Nsight-Cloud provides Operator, Injector, Coordinator, and Streamer components that make Nsight workflows usable in containerized clusters.

For development and debugging, use Nsight-Visual-Studio-Edition for Microsoft Visual Studio, Nsight-Visual-Studio-Code-Edition for VS Code, Nsight-Eclipse-Plugins for Eclipse-based CUDA workflows, CUDA-GDB for command-line CUDA debugging, and Compute-Sanitizer for functional correctness checking. Nsight-Integration links standalone Nsight tools back into the Visual Studio menu.

For notebook and Python workflows, use Nsight-JupyterLab-Extension to profile notebook cells with Nsight Systems or Nsight Compute, and Nsight-Python for Python-first profiling automation.

For APIs and SDKs, related pages include CUPTI for profiler/tracer construction, NVTX for application ranges and annotations, CUDA-Debugger-API for debugger integration, Nsight-Aftermath-SDK for graphics crash dump collection, and Nsight-Perf-SDK for in-application graphics performance metrics.

The current NVIDIA overview also lists legacy tools. Compute Sanitizer API coverage belongs in Compute-Sanitizer while NVIDIA keeps the API reference inside the current Compute Sanitizer documentation set.

Connections

Resources