67 packages
| Package ▲ | Version | Maintainers | # | Teams | Deps | Used by | Used by (transitive) | Status |
|---|---|---|---|---|---|---|---|---|
Building blocks that make it easier to write safe and efficient CUDA C++ code |
13.3.4.3.1 |
6 | 7 | 32 | 140 | team-only unfree | ||
Header-only library of GPU-accelerated, concurrent data structures |
0-unstable-2026-07-02 |
6 | 7 | 0 | 0 | maintained | ||
Samples for CUDA Developers which demonstrates features in CUDA Toolkit |
13.0 |
6 | 19 | 0 | 0 | maintained unfree | ||
Provides minor version forward compatibility for the CUDA runtime |
615.71.10 |
6 | 7 | 6 | 198 | team-only broken unfree | ||
13.4.92 |
6 | 7 | 9 | 156 | team-only unfree | |||
Analyzes trace files containing compilation time information generated by NVCC or NVRTC |
13.4.92 |
6 | 7 | 0 | 0 | team-only unfree | ||
CUDA Runtime |
13.4.92 |
6 | 12 | 107 | 135 | team-only unfree | ||
13.4.92 |
6 | 7 | 2 | 5 | team-only unfree | |||
Extracts information from CUDA binary files (both standalone and those embedded in host binaries) and presents them in human readable format |
13.4.92 |
6 | 7 | 3 | 26 | team-only unfree | ||
C-based interface for creating profiling and tracing tools designed for CUDA applications |
13.4.92 |
6 | 7 | 19 | 44 | team-only unfree | ||
Decode low-level identifiers that have been mangled by CUDA C++ into user readable names |
13.4.92 |
6 | 7 | 3 | 23 | team-only unfree | ||
13.4.92 |
6 | 7 | 0 | 0 | team-only unfree | |||
NVIDIA tool for debugging CUDA applications on Linux and QNX systems |
13.4.92 |
6 | 11 | 1 | 21 | team-only unfree | ||
CUDA compiler driver |
13.4.92 |
6 | 12 | 91 | 128 | team-only unfree | ||
Extracts information from standalone cubin files and presents them in human readable format |
13.4.92 |
6 | 7 | 3 | 26 | team-only unfree | ||
C-based programmatic interface for monitoring and managing various states within Data Center GPUs |
13.4.92 |
6 | 8 | 12 | 39 | team-only unfree | ||
Prune host object files and libraries to only contain device code for the specified targets |
13.4.92 |
6 | 7 | 1 | 21 | team-only unfree | ||
Runtime compilation library for CUDA C++ |
13.4.92 |
6 | 8 | 29 | 100 | team-only unfree | ||
Low-level API for heterogeneous computing that runs on CUDA-powered GPUs |
13.4.92 |
6 | 7 | 0 | 0 | team-only unfree | ||
API for profiling CUDA runtime |
13.4.92 |
6 | 7 | 19 | 53 | team-only unfree | ||
Enables the creation of sanitizing and tracing tools that target CUDA applications |
13.4.92 |
6 | 7 | 1 | 21 | team-only unfree | ||
CUDA Tile IR optimizing assembler |
13.4.92 |
6 | 9 | 2 | 6 | team-only unfree | ||
Wrapper substituting the deprecated runfile-based CUDA installation |
13.4 |
6 | 24 | 4 | 20 | team-only unfree | ||
GPU-accelerated library of primitives for deep neural networks |
9.27.0.42 |
7 | 10 | 21 | 30 | maintained unfree | ||
Python and C++ Graph API with SOTA attention (SDPA / Flash Attention), MoE grouped GEMM fusions, and FP8/MXFP8 kernels for Hopper and Blackwell GPUs |
1.32.2 |
6 | 10 | 4 | 4 | maintained | ||
Set of high-performance libraries and tools for accelerating quantum computing simulations at both the circuit and device level by orders of magnitude |
26.09.0.8 |
6 | 14 | 0 | 0 | team-only unfree | ||
CUDA Templates for Linear Algebra Subroutines |
4.8.1 |
6 | 12 | 3 | 3 | maintained | ||
GPU-accelerated vector search and clustering (CAGRA, IVF-Flat, IVF-PQ, brute force) |
26.08.01 |
6 | 18 | 0 | 0 | maintained | ||
GPU-accelerated library for data loading and pre-processing for deep learning |
2.4.0 |
6 | 27 | 2 | 2 | maintained | ||
Fast GPU memory copy library based on NVIDIA GPUDirect RDMA technology |
2.6 |
6 | 4 | 2 | 42 | maintained | ||
CUDA Basic Linear Algebra Subroutine library |
13.8.2.12 |
6 | 9 | 47 | 76 | team-only unfree | ||
High-performance, multi-process, GPU-accelerated library for distributed basic dense linear algebra |
0.11.0.4074 |
6 | 9 | 1 | 2 | team-only unfree | ||
Library of GPU-accelerated linear solvers with sparse matrices |
1.0.0.35 |
6 | 11 | 0 | 0 | team-only unfree | ||
High-performance FFT product CUDA library |
12.4.0.43 |
6 | 10 | 25 | 50 | team-only unfree | ||
Library to leverage GDS technology |
1.19.1.55 |
6 | 12 | 10 | 53 | team-only unfree | ||
CUDA cuObject Client |
1.3.1.55 |
6 | 10 | 0 | 0 | team-only unfree | ||
GPU-accelerated random number generation library |
10.4.4.72 |
6 | 8 | 29 | 53 | team-only unfree | ||
Collection of dense and sparse direct linear solvers and Eigen solvers |
12.3.5.3 |
6 | 11 | 28 | 53 | team-only unfree | ||
High-performance, distributed-memory, GPU-accelerated library that provides tools for solving dense linear systems and eigenvalue problems |
0.9.1.9318 |
6 | 12 | 1 | 1 | team-only unfree | ||
GPU-accelerated basic linear algebra subroutines for sparse matrix computations for unstructured sparsity |
12.8.6.72 |
6 | 9 | 27 | 56 | team-only unfree | ||
High-performance CUDA library dedicated to general matrix-matrix operations in which at least one operand is a structured sparse matrix with 50% sparsity ratio |
0.10.0.12 |
7 | 8 | 6 | 11 | maintained unfree | ||
GPU-accelerated tensor linear algebra library for tensor contraction, reduction, and elementwise operations |
2.8.1.0 |
6 | 10 | 3 | 3 | team-only unfree | ||
Library of primitives for image and signal processing |
13.2.0.58 |
6 | 8 | 5 | 25 | team-only unfree | ||
APIs which can be used at runtime to combine multiple CUDA objects into one CUDA fat binary (fatbin) |
13.4.92 |
6 | 8 | 0 | 0 | team-only unfree | ||
APIs which can be used at runtime to link together GPU device code |
13.4.92 |
6 | 8 | 16 | 68 | team-only unfree | ||
Provides high-performance, GPU accelerated JPEG decoding functionality for image formats commonly used in deep learning and hyperscale multimedia applications |
13.2.3.58 |
6 | 8 | 2 | 5 | team-only unfree | ||
Accelerates the decoding and encoding of JPEG2000 images on NVIDIA GPUs |
0.12.0.58 |
6 | 7 | 1 | 4 | team-only unfree | ||
APIs which can be used to compile a PTX program into GPU assembly code |
13.4.92 |
6 | 7 | 0 | 0 | team-only unfree | ||
Parallel programming interface for NVIDIA GPUs based on OpenSHMEM |
3.8.0-0 |
6 | 22 | 4 | 9 | maintained unfree | ||
Accelerates TIFF encode/decode on NVIDIA GPUs |
0.9.0.86 |
6 | 7 | 1 | 4 | team-only unfree | ||
Interface for generating PTX code from both binary and text NVVM IR inputs |
13.4.92 |
6 | 7 | 3 | 129 | team-only unfree | ||
NVIDIA Math Libraries Device Extensions (cuBLASDx, cuFFTDx, cuRANDDx, cuSolverDx, and nvCOMPDx) |
26.10.0 |
6 | 7 | 0 | 0 | team-only unfree | ||
Multi-GPU and multi-node collective communication primitives for NVIDIA GPUs |
2.32.3-1 |
7 | 11 | 31 | 41 | maintained | ||
Expert-parallelism (MoE dispatch/combine) extension library for NCCL |
2.32.3-1 |
1 | 9 | 0 | 0 | single | ||
RDMA and SHARP network plugins for NCCL |
2.7-unstable-2026-08-11 |
6 | 8 | 0 | 0 | maintained | ||
Tests to check both the performance and the correctness of NVIDIA NCCL operations |
2.21.1 |
7 | 8 | 0 | 0 | maintained | ||
Interactive profiler for CUDA and NVIDIA OptiX |
2026.3.1.2 |
6 | 13 | 0 | 0 | team-only unfree | ||
System-wide performance analysis and visualization tool |
2026.3.2.476 |
6 | 31 | 0 | 0 | team-only unfree | ||
Tool for bandwidth measurements on NVIDIA GPUs |
0.10.0 |
1 | 7 | 0 | 0 | single broken | ||
High-speed data compression and decompression library optimized for NVIDIA GPUs |
6.0.0.12 |
6 | 7 | 3 | 3 | team-only unfree | ||
Library of GPU- and CPU-accelerated image codecs with a unified interface |
0.10.0 |
6 | 17 | 3 | 3 | maintained | ||
C-based Application Programming Interface (API) for annotating events, code ranges, and resources in your applications |
3.7.0 |
6 | 4 | 20 | 45 | maintained | ||
RAPIDS Analytics Framework Toolkit for data science and ML |
26.08.00 |
6 | 13 | 1 | 1 | maintained | ||
RAPIDS Memory Manager |
26.08.00 |
6 | 9 | 1 | 2 | maintained | ||
Simple (Single-precision AX Plus Y) FindCUDAToolkit.cmake example for testing cross-compilation |
0-unstable-2023-07-11 |
6 | 7 | 0 | 0 | team-only | ||
SDK that facilitates high-performance machine learning inference |
11.4.0.106 |
6 | 9 | 5 | 7 | team-only unfree | ||
Open Source Software (OSS) components of NVIDIA TensorRT |
11.4.0 |
6 | 7 | 0 | 0 | maintained |