Boost Performance and Scale Artificial Intelligence and HPC Applications.

Within every company’s ever-growing volumes of data lie insights that can revolutionize their industry. Diagnosing cancer, predicting hurricanes, and automating business operations for improved efficiencies are just some of the breakthroughs you can accomplish with accelerated compute. AWS and NVIDIA have partnered to deliver the most powerful and advanced GPU-accelerated cloud to help clients build a more intelligent future.

Deep-learning tools save us tens of millions, if not hundreds of millions, of dollars over many years.

- Steve Phillpott, CIO, Western Digital

With Amazon EC2 P3 instances, we have the ability to run training workloads faster, enabling us to iterate more, build better machine learning models and reduce cost.

- Nick Handel, Head of ML, Airbnb

P3 instances gives our customers the ability to bring lifesaving drugs to market more quickly…P3 instances with their high-performance GPUs allow us to perform 4X as many simulations in a day as we could with P2 instances.

- Robert Abel, Senior Vice President of Science, Schrödinger

GPU-Accelerated Cloud Services

AMAZON EC2 P3

Amazon EC2 P3 instances are powered by NVIDIA V100 Tensor Core GPUs and Amazon infrastructure stack. Armed with eight of these instances interconnected for optimum scaling, customers can achieve up to 1 petaflop of breakthrough performance, innovate with any AI framework, and accelerate scientific simulations on over 550 top HPC applications. The NVIDIA suite pairs with Amazon higher stack platforms and APIs to give researchers, developers, and engineers the comprehensive toolset they need to address the toughest challenges. It also allows them to spend minimal time on setup or infrastructure management and realize tremendous cost savings as they scale their solutions.

EC2 P3 (NVIDIA® Tesla® V100)

Highest Performance for Compute

EC2 P3 (NVIDIA® Tesla® V100)

EC2 P2 (NVIDIA Tesla K80)

Entry-Level Compute

EC2 P2 (NVIDIA Tesla K80)

EC2 G3 (NVIDIA Tesla M60)

Graphic Visualization

EC2 G3 (NVIDIA Tesla M60)

Nvidia GPU Cloud (NGC)

Get simple access to a broad range of performance-engineered containers for AI, HPC, and HPC visualization to run on Amazon EC2 P3 instances from the NGC container registry. The containers from NGC include all necessary dependencies such as NVIDIA CUDA® runtime, NVIDIA libraries, and an operating system, and are tuned across the stack for optimal performance. To run NGC containers on AWS and take full advantage of the NVIDIA Volta architecture, NVIDIA developed the NVIDIA Volta Deep Learning AMI, available from the AWS Marketplace.

GET STARTED WITH NGC

Apache MXNET

Apache MXNet is a fast, scalable training and inference deep learning framework. It includes the Gluon interface to let developers of all skill levels get started with deep learning on the cloud, edge devices, and mobile apps. Working with NVIDIA, AWS engineers and researchers have pre-optimized neural machine translation (NMT) algorithms on Apache MXNet. This approach allows developers and data scientists to train faster than ever before on Volta-based platforms.

Sign up for an AWS account to get instant access to the AWS Free Tier and start building with MXNet.

GPU-Accelerated Virtualized Graphics

NVIDIA GRID® for GPU-accelerated graphics helps creative and technical professionals maximize their productivity from anywhere by accessing the most demanding professional design and engineering applications from the cloud. Amazon Web Service gives designers and engineers the flexibility to run virtual desktops on EC2 P3 instance powered by NVIDIA Tesla V100 GPUs.

Access the Power of AWS and Volta

Amazon EC2 P3 Instances

Optimized Containers from NGC