Software Solutions for
Deep Learning & AI


To begin a deep learning project you need some understanding of the relevant programming languages and techniques required. However, to avoid learning an entire language, there are software applications, deep learning frameworks and libraries that have been designed to automate and structure some of the tasks involved. The majority of these applications and frameworks are available for free from the NVIDIA GPU Cloud (NGC).

What is the NVIDIA GPU Cloud?

The NVIDIA GPU Cloud (NGC) is an online repository of GPU-accelerated software for deep learning, machine learning and high-performance computing (HPC) that takes care of all the plumbing so data scientists, developers, and researchers can focus on building solutions, gathering insights, and delivering business value. NGC containers deliver powerful and easy-to-deploy software proven to deliver the fastest results.

The NGC catalogue accelerates productivity with easy-to-deploy, optimised AI and HPC application containers, so users can take care of the heavy lifting (expertise, time, compute resources) with pre-trained models and workflows with best-in-class accuracy and performance.

Applications and frameworks from the NGC catalogue can be run on-premise, in the cloud or at the edge - either on bare metal servers, containerised or virtualised environments - maximising GPU utilisation and the portability and scalability of applications.

Enterprise-grade support is provided for NVIDIA-Certified platforms supporting DGX, HGX, EGX, RTX and AGX systems. This provides direct access to NVIDIA's AI experts, minimising risk and downtime, while maximising system efficiency and productivity.

Applications or Frameworks

The building blocks of all AI software applications and deep learning frameworks are the underlying programming languages and their libraries. Applications automate as much programming as possible, so you only need to customise and add to the groundwork already in place to deliver your desired outcome. Frameworks require much more programming knowledge and input, but allow for a more specific and tailored result. The two sections below cover common application uses and outcomes; and the most popular frameworks and libraries.

This collection of applications - the majority provided free-of-charge via NGC - delivers pre-trained and configured frameworks, libraries and content designed to accelerate AI research for specific use cases. Click the categories to learn more.

These healthcare and life sciences applications are designed for researchers looking to train AI models for medical imaging, genomics, patient monitoring and drug discovery.


AI Accelerated Healthcare

NVIDIA Clara is family of healthcare applications aimed at AI-powered imaging, genomics, and the development and deployment of smart sensors. It includes full-stack GPU-accelerated libraries, SDKs, and reference applications for developers, data scientists, and researchers to create real-time, secure and scalable solutions.


Medical Devices

Clara Holoscan is an AI computing platform for medical devices that combines hardware systems for low-latency sensor and network connectivity. It also optimises libraries for data processing and AI, and accelerates core micro-services for streaming, imaging, and other applications—from embedded to edge to cloud.



The Clara Parabricks application provides both enterprise-grade, turnkey, GPU-accelerated sequencing software and a technology stack for developers to build applications for high-performance computing, deep learning and data analytics in genomics.


Drug Discovery

Clara Discovery is a collection of frameworks, applications, and AI models that, together, accelerate drug discovery, supporting research in genomics, microscopy, virtual screening, computational chemistry, visualisation, clinical imaging and more.


Smart Hospitals

Clara Guardian is an application framework that brings video analytics and conversational AI capabilities to hospitals, simplifying the development and deployment of smart sensors to enhance clinical experiences


Medical Imaging

The Clara Imaging application accelerates medical imaging AI workflows with open-source frameworks, AI-assisted annotation, AI inference and pre-trained models.

These audio and video applications are designed deliver automated speech recognition, text-to-speech, speech-to-text or speech-to-speech translation, natural language programming (NLP) for uses such as virtual assistants, animated characters, video-conferencing and virtual collaboration.


Video Conferencing

NVIDIA Maxine is a suite of GPU-accelerated SDKs that reinvent audio and video communications with AI, elevating standard microphones and cameras for clear online communications. Maxine provides state-of-the-art real-time AI audio, video and augmented reality features that can be built into customisable, end-to-end deep learning pipelines.


Speech Recognition

NVIDIA Riva is aimed at building Speech AI applications that are customised for your use case and deliver real-time applications such as virtual assistants, call centre agent assist and video conferencing. Riva components are customisable, so you can adapt the applications for your use case and industry and deploy them in any cloud, on-premises and at the edge.


Conversational AI

NVIDIA NeMo is an open-source application for developers to build, train and fine-tune GPU-accelerated speech models for real-time automated speech recognition (ASR), natural language processing (NLP), and text-to-speech (TTS) applications such as video call transcriptions, intelligent video assistants, and automated call centre support across healthcare, finance, retail and telecommunications


Character Animation

The NVIDIA Avatar application enables developers and artists to generate, animate, simulate and render state-of-the-art interactive avatars as well as the experiences that use them. Key uses are in game design, virtual assistants and video communication


Real Time Collaboration

NVIDIA Omniverse is a collaboration and scalable multi-GPU, real-time, true-to-reality simulation platform. Omniverse revolutionises the way you can create and develop as individuals and work together as teams, bringing more creative possibilities and efficiency to 3D creators and developers.

These applications are designed to accelerate and enhance robotics research - from development to simulation, through to deployment.


Robotic Development

The NVIDIA Isaac suite of applications aids building and deploying commercial-grade, AI-powered robots. Isaac is a toolkit that includes building blocks and tools that accelerate robot developments such as GPU-accelerated algorithms and deep neural networks (DNNs) for perception and planning, plus machine learning workflows for supervised and reinforcement learning.


Autonomous Robots

The Isaac for Autonomous Mobile Robots (AMR) platform extends NVIDIA Isaac capabilities for developers building and deploying robotics applications by bringing mapping, site analytics and fleet and route optimisation onto NVIDIA EGX servers. This platform helps enhance and accelerate AMR applications for the logistics industry in applications ranging from warehouses to retail.


Robotic Simulation

Isaac Sim provides developers with a data cockpit to synthetically generate datasets for machine learning (ML) models from easy-to-understand parameters. Deployed within Omniverse, Isaac Sim generates synthetic data that can be used to train DNNs running on an AMR. This helps developers build and deploy AI-enabled robots that operate safely and avoid common mishaps.

These applications are designed to aid the development of datasets and their visualisation, recommendation systems and to help the understanding of researchers, engineers and designers when it comes to physics-driven modelling such as fluid dynamics, modular dynamics, climate models and mechanics.


Data Preparation

Driverless AI by is a machine learning platform that automates many of the most difficult data science and machine learning workflows, such as feature engineering, model validation, model tuning, model selection and model deployment. It enables the rapid development of hundreds of machine learning models to help your business mitigate risks and maximise revenue potential.


Physics Simulations

NVIDIA Modulus is a neural network framework that blends the power of physics in the form of governing partial differential equations (PDEs) with data to build high-fidelity, parameterised surrogate models with near-real-time latency. Whether you’re looking to get started with AI-driven physics problems or designing digital twin models for complex non-linear, multi-physics systems, NVIDIA Modulus can be used independently or within Omniverse.



NVIDIA Merlin is an open-source application for building high-performing recommender systems at scale. It includes libraries, methods and tools that streamline the building of recommenders by addressing common preprocessing, feature engineering, training and inference challenges. Each component of the Merlin pipeline is optimised to support hundreds of terabytes of data, all accessible through easy-to-use APIs.

These applications are designed to aid the development of smart and automated infrastructure projects including smart cities, factories, airports and autonomous vehicles.


Smart Cities

The NVIDIA Metropolis application is an end-to-end application framework that combines pre-trained models, training and optimisation tools, with common video cameras and sensors with AI-enabled video analytics to provide operational efficiency and safety applications across a broad range of industries - including retail analytics, city traffic management, airport operations and automated factory


Autonomous Vehicles

The NVIDIA DRIVE application for the development of autonomous vehicles covers everything from the car to the datacentre. It includes highly automated supervised driving and an AI cockpit, empowering developers to efficiently build and deploy a variety of state-of-art features, including perception, localisation and mapping, planning and control, driver monitoring and natural language processing.


Driving Simulation

NVIDIA DRIVE Sim uses high-fidelity and physically accurate simulation to create a safe, scalable and cost-effective way to deploy self-driving vehicles. It delivers a powerful, cloud-based computing platform capable of generating a wide range of real-world scenarios for AV development and validation, in either a standalone environment or deployed within Omniverse.

These applications are designed to ensure maximum utilisation of GPU hardware. They work by allowing the sharing of GPU resource so that they can be allocated and segregated across users and tasks as demand requires.


Sharing a GPU

Multi-Instance GPU (MIG) expands the performance and value of NVIDIA H100, A100, and A30 Tensor Core GPUs. MIG can partition the GPU into as many as seven instances, each fully isolated with its own high-bandwidth memory, cache, and compute cores. This gives administrators the ability to support every workload, from the smallest to the largest, with guaranteed quality of service (QoS) and extending the reach of accelerated computing resources to every user.  


Pooling GPUs

The Run:ai Atlas software platform decouples data science workloads from the underlying hardware - regardless of what hardware you have. By pooling resources and applying an advanced scheduling mechanism to data science workflows, Run:ai greatly increases the ability to fully utilise all available resources, essentially creating unlimited compute. Data scientists can increase the number of experiments they run, speed time to results and ultimately meet the business goals of their AI initiatives.

Find out more

When you require greater control over your AI application development, it may be better to start with frameworks, libraries or even programming languages if you possess the knowledge. These can all be tailored and altered to greater or lesser degrees, as required, as opposed to dedicated pre-configured applications that are designed for a single use case.

A framework is essentially code written by someone else that helps you perform some common tasks in a less verbose way. A framework inverts the control of the program and tells the developer what they need.

Framework Features
PyTorch is an open source machine learning framework based on the Torch library, used for applications such as computer vision and natural language processing. Caffe2 is now integrated into PyTorch.
TensorFlow is an end-to-end open source platform for machine learning. It has a comprehensive, flexible ecosystem of tools, libraries and community resources that lets researchers push the state-of-the-art in ML and developers easily build and deploy ML powered applications.
The Microsoft Cognitive Toolkit (CNTK) is an open-source toolkit for commercial-grade distributed deep learning. It describes neural networks as a series of computational steps via a directed graph. CNTK allows the user to easily realize and combine popular model types such as feed-forward DNNs, convolutional neural networks (CNNs) and recurrent neural networks (RNNs/LSTMs). CNTK implements stochastic gradient descent (SGD, error backpropagation) learning with automatic differentiation and parallelization across multiple GPUs and servers.
Chainer is an open source deep learning framework written purely in Python on top of NumPy and CuPy Python libraries. The development is led by Japanese venture company Preferred Networks in partnership with IBM, Intel, Microsoft, and Nvidia.
Apache MXNet is an open-source deep learning software framework, used to train, and deploy deep neural networks. It is scalable, allowing for fast model training, and supports a flexible programming model and multiple programming languages.
Caffe is developed by the Berkeley Vision and Learning Center (BVLC) and by community contributors. NVIDIA Caffe, also known as NVCaffe, is a NVIDIA-Maintained fork of BVLC Caffe tuned for NVIDIA GPUs, particularly in multi-GPU configurations.

A library is similar to a framework, the major difference being that the developer is in charge of the code rather than being offered pointers, as a framework does.

Framework Features
Keras is an open-source software library that provides a Python interface for artificial neural networks. Keras acts as an interface for the TensorFlow library. Up until version 2.3, Keras supported multiple backends, including TensorFlow, Microsoft Cognitive Toolkit, Theano, and PlaidML.
Theano is a Python library and optimizing compiler for manipulating and evaluating mathematical expressions, especially matrix-valued ones. In Theano, computations are expressed using a NumPy-esque syntax and compiled to run efficiently on either CPU or GPU architectures.
Torch is an open-source machine learning library, a scientific computing framework, and a script language based on the Lua programming language. It provides a wide range of algorithms for deep learning, and uses the scripting language LuaJIT and an underlying C implementation.
RAPIDS is a suite of open-source software libraries and APIs for executing data science pipelines entirely on GPUs—and can reduce training times from days to minutes. Built on NVIDIA CUDA-X AI, RAPIDS unites years of development in graphics, machine learning, deep learning, high-performance computing (HPC) and more.
Gluon is an open source deep learning library jointly created by AWS and Microsoft that helps developers build, train and deploy machine learning models in the cloud.

Using a programming language from scratch offers full control of the AI model creation process. This requires extensive knowledge although Python is the most popular programming language due to its syntaxes being very simple and can be easily learnt, which makes algorithms easier to implement.

Framework Features
Python is a high-level, general-purpose programming language. Its design philosophy emphasises code readability with the use of significant indentation. Its language constructs and object-oriented approach aim to help programmers write clear, logical code for small- and large-scale projects.
MATLAB is a proprietary multi-paradigm programming language and numeric computing environment developed by MathWorks. MATLAB allows matrix manipulations, plotting of functions and data, implementation of algorithms, creation of user interfaces, and interfacing with programs written in other languages.
R is a programming language for statistical computing and graphics supported by the R Core Team and the R Foundation for Statistical Computing. Created by statisticians, R is used among data miners and statisticians for data analysis and developing statistical software.
Julia is a high-level, high-performance, dynamic programming language. While it is a general-purpose language and can be used to write any application, many of its features are well suited for numerical analysis and computational science.
C++ is a popular programming language. C++ is used to create computer programs, and is one of the most used language in game development.
JavaScript, often abbreviated JS, is a programming language that is one of the core technologies of the World Wide Web, alongside HTML and CSS. Over 97% of websites use JavaScript on the client side for web page behaviour, often incorporating third-party libraries.

An orchestration tool designed to make it easier to create, deploy, and run projects side by side by using containers. Containers allow a developer to package up a workflow with all of the parts it needs, such as data, libraries and other dependencies and deploy it as one package. Containers also allow multiple models to be deployed but in segregation from one another. This also means that any changes or optimisations made within a container will not impact the host operating system or other containers.

Framework Features
Docker is a suite of software development tools for creating, sharing and running individual containers on a single node.
Kubernetes is a system for operating containerised applications at scale, capable of running them across clusters.

Discover the Scan AI range of development, training and
inferencing hardware solutions.

Find out more
Find out more Find out more