nvidia-nvtx-cu11

What it is and what it does

nvidia-nvtx-cu11 is NVIDIA's Tools Extension library for CUDA 11, providing a C-based API to instrument applications with event markers, code range annotations, and resource labels. These annotations are captured by NVIDIA's Visual Profiler to visualize performance characteristics and identify bottlenecks during GPU-accelerated workloads.

The package targets machine learning, deep learning, and scientific computing workflows where profiling GPU execution is critical. It has no runtime dependencies and ships as platform-specific wheels for Linux (x86_64 and aarch64) and Windows. However, the package has been abandoned for 1411 days with no recent updates, raising concerns about compatibility with newer CUDA versions and modern profiling tools.

Use it for:

• Annotate GPU kernel launches and memory transfers in CUDA applications for Visual Profiler visualization
• Mark critical code sections in deep learning training loops to identify performance bottlenecks
• Label custom CUDA kernels and resource allocations for detailed performance analysis
• Instrument scientific computing applications to correlate GPU events with CPU-side code execution

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Provides a C-based API for annotating events, code ranges, and resources in applications to enable capture and visualization through NVIDIA's Visual Profiler.

No—the package is abandoned (1411 days since last release) with unclear proprietary licensing. While it provides profiling instrumentation for CUDA 11 workloads, lack of maintenance means no support for newer CUDA versions or Python releases. Consider alternative profiling approaches instead.

Install

nvidia-nvtx-cu11 on PyPI

pip install nvidia-nvtx-cu11

uv add nvidia-nvtx-cu11

poetry add nvidia-nvtx-cu11

Installing nvidia-nvtx-cu11

Before you install

Medium install friction due to platform-specific wheels (x86_64 Linux, aarch64, Windows). Package is abandoned as of 1411 days since last release, with no recent maintenance visible.

License in practice

Licensed under NVIDIA Proprietary Software with unclear treatment—not an open-source license. Verify licensing terms with NVIDIA before use in commercial or redistributed applications.

Quickstart

pip install nvidia-nvtx-cu11
import nvidia_nvtx_cu11
# Use NVTX API to annotate events for Visual Profiler capture

Requires CUDA 11 environment and compatible GPU; platform-specific wheels limit portability to x86_64 Linux, aarch64, or Windows.

Verify before relying

• Whether NVTX annotations work with CUDA versions beyond 11.x
• Compatibility with Python versions beyond 3.11 despite 'supports_current' classifier
• Whether abandoned status affects Visual Profiler integration in recent NVIDIA tools

Package facts

Evidence: nvidia_nvtx_cu11-11.8.86-py3-none-manylinux1_x86_64.whl; nvidia_nvtx_cu11-11.8.86-py3-none-manylinux2014_aarch64.whl; nvidia_nvtx_cu11-11.8.86-py3-none-manylinux2014_x86_64.whl; nvidia_nvtx_cu11-11.8.86-py3-none-win_amd64.whl

Keywords: cuda, nvidia, runtime, machine learning, deep learning

Tags

More Software Development packages

Further reading