The NVIDIA CUDA® Deep Neural Network (cuDNN) is a GPU-accelerated library of primitives for deep neural networks. cuDNN provides highly tuned implementations for standard routines such as forward and backward convolution, attention, matmul, pooling, and normalization.


These documents provide information regarding the current NVIDIA cuDNN 8.9.5 release.

These release notes describe the key features, software enhancements and improvements, and known issues.
Provides the supported versions of the OS, NVIDIA CUDA, the CUDA driver, and the hardware.
Provides step-by-step instructions on how to install and check for correct operation of NVIDIA cuDNN on Linux and Microsoft Windows systems.
Provides the datatypes and functions per library. Specifically, this reference consists of a cuDNN datatype reference section that describes the types of enums and a cuDNN API reference section that describes all routines in the cuDNN library API. The cuDNN API is a context-based API that allows for easy multithreading and (optional) interoperability with CUDA streams.
Provides step-by-step instructions on how to use the NVIDIA cuDNN library. While the NVIDIA cuDNN API Reference provides per-function API documentation, the Developer Guide gives a more informal end-to-end story about cuDNN’s key capabilities and how to use them.

These archives provide access to previously released cuDNN documentation versions.