nvidia-cusparselt-cu13
NVIDIA cuSPARSELt
Decision gist · record as of 2026-08-14
Yes, if you are developing GPU-accelerated applications on NVIDIA hardware with structured sparse matrices and have CUDA 13 available. The library is actively maintained, has no known vulnerabilities, and ranks in the top 1000 PyPI packages. However, verify NVIDIA's proprietary license terms for your use case, confirm your GPU architecture is supported (SM 8.0+), and ensure CUDA 13 is installed and compatible with your system before proceeding.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires CUDA 13 runtime, compatible NVIDIA GPU (SM architecture 8.0+), and platform support (Linux aarch64/x86_64 or Windows x86_64).
- Medium install friction due to platform-specific wheels (aarch64, x86_64, Windows only); requires CUDA 13 and compatible NVIDIA GPU hardware.
- Package is actively maintained with recent releases.
License · maintenance · safety
NVIDIA Proprietary Software (unclear) — Licensed under NVIDIA Proprietary Software with unclear treatment terms; users should verify licensing compliance with NVIDIA before deployment in production or commercial contexts.
last release 2026-04-29 (107 days)
0 known vulnerabilities (OSV.dev, 2026-08-14) · 42,690,252 downloads/mo, #654 on PyPI
Alternatives
Verify before relying
pip install nvidia-cusparselt-cu13
import nvidia.cusparselt as cusparselt
# Requires CUDA 13 runtime and compatible NVIDIA GPU- Whether Python bindings are included or if this is a C/C++ library wrapper requiring additional setup
- Exact Python version compatibility and whether requires_python constraint exists
- Whether documentation or examples are bundled or require separate download from developer.nvidia.com
What it is and what it does
nvidia-cusparselt-cu13 is NVIDIA's CUDA library for accelerating sparse matrix-matrix multiplication on GPUs. It targets operations where at least one operand is a structured sparse matrix with 50% sparsity ratio, exploiting NVIDIA's Sparse MMA tensor cores for performance. The library supports flexible algorithm selection, mixed-precision computation (FP32, BF16, FP16, INT8, E4M3, E5M2), and various matrix characteristics including memory layout and alignment.
This is a low-level GPU compute library, not a high-level Python framework. It binds to CUDA 13 and requires compatible NVIDIA hardware (SM architecture 8.0 and above). Installation is platform-specific, with wheels provided only for Linux aarch64, Linux x86_64, and Windows x86_64. The package is actively maintained and carries no known security vulnerabilities.
Use it for
- Accelerate sparse matrix operations in machine learning inference pipelines on NVIDIA GPUs with structured sparsity patterns.
- Optimize matrix-matrix multiplication in high-performance computing applications that exploit Ampere or newer GPU tensor cores.
- Implement mixed-precision sparse linear algebra in deep learning frameworks that support structured sparsity.
- Reduce memory bandwidth and latency in sparse neural network computations by leveraging GPU sparse tensor cores.
- Build custom CUDA kernels for scientific computing that require efficient sparse matrix operations on NVIDIA hardware.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you are developing GPU-accelerated applications on NVIDIA hardware with structured sparse matrices and have CUDA 13 available.
The library is actively maintained, has no known vulnerabilities, and ranks in the top 1000 PyPI packages. However, verify NVIDIA's proprietary license terms for your use case, confirm your GPU architecture is supported (SM 8.0+), and ensure CUDA 13 is installed and compatible with your system before proceeding.
Install
nvidia-cusparselt-cu13 on PyPI
Before you install
Medium install friction due to platform-specific wheels (aarch64, x86_64, Windows only); requires CUDA 13 and compatible NVIDIA GPU hardware. Package is actively maintained with recent releases.
Requires CUDA 13 runtime, compatible NVIDIA GPU (SM architecture 8.0+), and platform support (Linux aarch64/x86_64 or Windows x86_64).
License in practice
Licensed under NVIDIA Proprietary Software with unclear treatment terms; users should verify licensing compliance with NVIDIA before deployment in production or commercial contexts.
Quickstart
pip install nvidia-cusparselt-cu13
import nvidia.cusparselt as cusparselt
# Requires CUDA 13 runtime and compatible NVIDIA GPUVerify before relying
- Whether Python bindings are included or if this is a C/C++ library wrapper requiring additional setup
- Exact Python version compatibility and whether requires_python constraint exists
- Whether documentation or examples are bundled or require separate download from developer.nvidia.com
Package facts
| License | NVIDIA Proprietary Software unclear |
| Python support | Not specified |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | None |
| Maintenance | Actively maintained 107 days since the last release |
| First released | |
| Downloads | 42,690,252 / month, #654 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Environment :: GPU :: NVIDIA CUDAEnvironment :: GPU :: NVIDIA CUDA :: 13Topic :: Scientific/Engineering |
Evidence: nvidia_cusparselt_cu13-0.9.1-py3-none-manylinux2014_aarch64.whl; nvidia_cusparselt_cu13-0.9.1-py3-none-manylinux2014_x86_64.whl; nvidia_cusparselt_cu13-0.9.1-py3-none-win_amd64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “sparse matrix multiplication cuda”
- nvidia-cusparselt-cu13Provides NVIDIA's CUDA library for high-performance sparse…
- nvidia-cusparselt-cu12Provides NVIDIA's cuSPARSELt CUDA library for high-performance sparse…
- nvidia-cusparse-cu11Provides CUSPARSE native runtime libraries for CUDA 11, enabling…
Give your agent the search over MCP, or paste the wish link into any chat.
More Scientific/Engineering packages
NumPy provides an N-dimensional array object and a comprehensive suite of mathematical, linear algebra, Fourier transform, and random number functions for scientific computing in Python.
pandas provides fast, flexible data structures (Series and DataFrame) for loading, cleaning, transforming, and analyzing labeled or relational data in Python.
scipy provides numerical algorithms for mathematics, science, and engineering—including optimization, integration, linear algebra, Fourier transforms, signal and image processing, and ODE solvers—built on numpy arrays.
scikit-learn provides a comprehensive Python library for supervised and unsupervised machine learning, including classification, regression, clustering, dimensionality reduction, and model evaluation tools built on NumPy and SciPy.
Install it if you need to train, evaluate, or deploy supervised or unsupervised learning models.
dill extends Python's pickle module to serialize and deserialize a much wider range of Python objects, including functions, lambdas, classes, and interpreter sessions, to byte streams for storage or network transmission.
Multiprocess is an enhanced fork of Python's standard multiprocessing library that uses dill for better serialization, allowing you to spawn processes with a threading-like API and share complex objects between them.
Install it if you use multiprocessing and encounter pickle serialization limits with lambdas or complex objects.
See also nvidia-cusparselt-cu12 · sparse-dot-topn · humming-kernels · cutensor-cu13 · qdldl · cutensor-cu12 · nvidia-cusparse-cu12 · nvidia-cusparse · nvidia-cusparse-cu11 · causal-conv1d