GPU-accelerated cloud images from NVIDIA® enable researchers, data scientists, and developers to harness the power of GPU computing in the cloud and on-demand. With preconfigured virtual images and containers loaded with drivers, the NVIDIA CUDA® Toolkit and deep learning software, data scientists and developers can get started accelerating their applications in minutes.

To get started, choose from the preconfigured GPU cloud images below or sign-up for NVIDIA GPU CLOUD to get access to NVIDIA optimized deep learning framework containers including NVCaffe, Caffe2, Microsoft Cognitive Toolkit (CNTK), Digits, MXNet, PyTorch, TensorFlow, Theano, and Torch.


Sign up for NGC


NVIDIA DIGITS 6 AMI

NVIDIA maintained Amazon Machine Image (AMI) with NVIDIA® DIGITS™ on Ubuntu operating system.

DIGITS puts the power of deep learning in the hands of data scientists and researchers. It simplifies common deep learning tasks such as managing data, designing and training neural networks, monitoring performance and choosing the best model.

Using DIGITS data scientists can rapidly design the best deep neural network (DNN) for image classification, segmentation and object detection tasks. Visit the DIGITS product page to learn more.

  1. Click Get Started button below to navigate to the DIGITS AMI on AWS marketplace using the button below
  2. Click continue
  3. Under 1-click, select closest geographical region
  4. Choose instance type depending on training performance needs (e.g. p2.8xlarge)
  5. Open a browser and type the public IP address of the instance at port 34448 (e.g. http://10.1.2.3:34448/) to get the DIGITS UI
  6. Follow these steps to log into the instance via SSH to download datasets
  7. To download the MNIST dataset for example, run the following commands on the console

    # mkdir $HOME/mnist
    # /usr/share/digits/tools/download_data/main.py mnist $HOME/mnist


Get Started


NVIDIA CUDA Toolkit 7.5 AMI

NVIDIA maintained Amazon Machine Image (AMI) with CUDA® Toolkit 7.5 on Amazon Linux 2016.03 (64-bit architecture) operating system.

The CUDA Toolkit provides a development environment for C/C++ developers building GPU-accelerated applications. It includes a compiler for NVIDIA GPUs, math libraries, and tools for debugging and optimizing the performance of your applications.

To get started with developing GPU-accelerated applications, you will find programming guides, user manuals, samples, and API references in the toolkit. Use the CUDA AMI to prototype, test and deploy your algorithms on single and multi-GPU configurations.

  1. Click the Get Started button below to navigate to NVIDIA CUDA Toolkit 7.5 AMI on AWS marketplace
  2. Click continue
  3. Under 1-click, select closest geographical region
  4. Choose instance type depending on workload needs
  5. Once the instance is running, use a terminal emulator (e.g. PuTTY on Windows) to connect to the instance via SSH


Get Started


Windows AMI with the NVIDIA Driver

Amazon EC2 running Microsoft Windows Server is a fast and dependable environment for deploying applications using the Microsoft Web Platform. This AMI based on Windows Server 2012 R2, comes installed with the latest NVIDIA driver and allows developers to develop and run CUDA applications on AWS's high-performance, reliable, cost-effective, cloud computing platform. Simply download and install the CUDA Toolkit to get started with developing GPU-accelerated applications.

  1. Click the Get Started button below to navigate to Windows AMI on AWS marketplace
  2. Click continue
  3. Under 1-click, select closest geographical region
  4. Choose instance type depending on workload needs
  5. Once the instance is running, right-click to obtain an auto-generated password for the Windows instance
  6. Use a program that supports RDP such as Remote Desktop Connection to connect to the Windows instance using the password
  7. Download and install the CUDA Toolkit and a development environment such as Visual Studio to get started with developing CUDA applications


Get Started