GPU-accelerated microservices and libraries for AI.

Developers, researchers, and inventors across a wide range of domains use GPU programming to accelerate their applications. Developing these applications requires a robust programming environment with highly optimized, domain-specific microservices and libraries. NVIDIA CUDA-X, built on top of CUDA®, is a collection of microservices, libraries, tools, and technologies for building applications that deliver dramatically higher performance than alternatives across data processing, AI, and high performance computing (HPC).

CUDA-X Microservices

Built by CUDA experts at NVIDIA, CUDA-X microservices are developer tools, GPU-accelerated libraries, and technologies packaged as cloud APIs. They are easy to integrate, customize, and deploy in data processing, AI, and HPC applications. 

CUDA-X microservices include NVIDIA® Riva for customizable speech and translation AI, NVIDIA Earth-2 for high-resolution climate and weather simulations, NVIDIA cuOpt™ for routing optimization and NVIDIA NeMo™ Retriever for responsive retrieval-augmented generation (RAG) capabilities for enterprises.

CUDA-X Libraries

CUDA-X Libraries are built on top of CUDA to simplify adoption of NVIDIA’s acceleration platform across data processing, AI, and HPC. With over 400 libraries, developers can easily build, optimize, deploy, and scale applications across PCs, workstations, the cloud, and supercomputers using the CUDA platform.


To build transformative AI applications, enterprises need to train models on their unique, proprietary data, at a time when datasets are growing by zettabytes every year. This poses an enormous compute problem that the CUDA-X data processing platform is designed to solve, using a suite of accelerated libraries that speed up and scale out processing of image, text, and tabular data.

Modern AI has the potential to disrupt many industries, but harnessing its power is challenging. Developing AI applications takes many steps—data processing, feature engineering, machine learning, verification, and deployment—and each step involves processing large volumes of data and performing massive computing operations. CUDA-X AI provides the tools and technologies needed to conquer this challenge.

HPC applications span many domains from fluid dynamics to weather simulation. CUDA-X HPC is a collection of libraries, tools, compilers and APIs that help developers solve the world's most challenging problems. CUDA-X HPC includes highly tuned kernels essential for high-performance computing (HPC). GPU-accelerated libraries for linear algebra, parallel algorithms, signal and image processing lay the foundation for compute-intensive applications in areas such as computational physics, chemistry, molecular dynamics, and seismic exploration.

Available Everywhere

CUDA-X is widely available. Its software-acceleration libraries are part of leading cloud platforms, including AWS, Microsoft Azure, and Google Cloud. They’re free as individual downloads or containerized software stacks from NGC. CUDA-X libraries can be deployed everywhere on NVIDIA GPUs, including desktops, workstations, servers, supercomputers, cloud computing, and internet of things (IoT) devices.

Over one million developers are using CUDA-X, providing the power to increase productivity while benefiting from continuous application performance. Whether you’re creating a new application or trying to speed up an existing one, CUDA-X provides the most efficient, effective path forward.

Unlock GPU Power for Applications

Explore the domains being transformed by NVIDIA CUDA-X and the GPU-accelerated libraries available within them.