Build, tune, and deploy accelerated applications with CUDA

Find installation instructions, launch highlights, programming guides, compiler documentation, API references, CUDA libraries, profiling tools, samples, and release notes for the CUDA Toolkit.

Install CUDA Release Notes Programming Guide

What's New in CUDA 13.4

Start with these CUDA 13.4 highlights, then open the release notes for the complete list of changes.

• Windows on Arm - The CUDA Toolkit is now supported on Windows on Arm platform for RTX Spark devices.
• NVIDIA Rubin preview - Adds preview support for the NVIDIA Rubin architecture.
• MPS V3 - Introduces CUDA MPS V3, simplifying the automation and management of shared GPU resources.

Read the CUDA Toolkit Release Notes for the complete CUDA 13.4 feature list, resolved issues, known issues, deprecations, and compatibility notes.

Explore CUDA documentation

Find CUDA documentation by workflow, feature area, or API/reference type.

Release and installation

• Release Notes
• Quick Start Guide
• Installation Guide Linux
• Installation Guide Windows
• CUDA Toolkit Archive

Learning and tutorials

• Accelerated Computing Hub
• CUDA Samples

Programming

• CUDA Programming Guide
• CUDA Best Practices Guide
• cuTile Python
• CUDA Tile IR
• CUDA Demo Suite
• CUDA Library Samples

CUDA Compiler

• NVCC
• NVRTC (Runtime Compilation)
• nvJitLink
• nvFatbin
• CompileIQ

PTX and low-level programming

• PTX ISA
• Inline PTX Assembly
• PTX Interoperability
• PTX Compiler APIs

CUDA APIs

• CUDA Runtime API
• CUDA Driver API
• CUDA Math API
• CUDA Tile C++ API Reference
• cuDLA API
• cuFile API Reference Guide
• CUDA Debugger API

CUDA libraries

• cuBLAS
• cuFFT
• cuRAND
• cuSOLVER
• cuSPARSE
• CUB
• CUDA C++ Standard Library
• Thrust

Additional libraries

• NPP
• NVBLAS
• nvJPEG

Architecture compatibility

• Blackwell Compatibility Guide
• Hopper Compatibility Guide
• Ada Compatibility Guide
• NVIDIA Ampere GPU Architecture Compatibility Guide
• Turing Compatibility Guide
• CUDA Compatibility

Architecture tuning

• Blackwell Tuning Guide
• Hopper Tuning Guide
• Ada Tuning Guide
• NVIDIA Ampere GPU Architecture Tuning Guide
• Turing Tuning Guide

Debugging and profiling

• Compute Sanitizer
• CUDA-GDB
• CUDA Compile Time Advisor
• CUPTI
• Nsight Compute
• Nsight Systems
• Nsight Visual Studio Edition
• CUDA Binary Utilities

GPUDirect and platforms

• GPUDirect RDMA
• GPUDirect Storage
• CUDA on WSL
• CUDA on EFLOW
• Multi-Instance GPU (MIG)
• vGPU

Compiler SDK

• libNVVM API
• libdevice User's Guide
• NVVM IR

Application notes and references

• CUDA for Tegra
• Video Decoder
• Floating Point and IEEE 754
• Incomplete-LU and Cholesky Preconditioned Iterative Methods
• CUDA Features Archive
• EULA