What are AWS Deep Learning AMIs?
AWS Deep Learning AMIs (DLAMI) provide customized machine images for deep learning in the cloud. They are available in most AWS Regions and for a wide range of Amazon EC2 instance types, from a small CPU-only instance to the latest high-powered multi-GPU instances.
The images come preconfigured with NVIDIA CUDA and NVIDIA cuDNN and the latest releases of the most popular deep learning frameworks. You launch a ready-to-use training or inference environment directly, without assembling drivers, libraries and frameworks yourself.
Core capabilities
- Preconfigured environment: NVIDIA CUDA, NVIDIA cuDNN and recent deep learning frameworks already installed
- Selection by type: Base, Single-Framework and Multi-Framework (Conda DLAMI)
- Architectures: x86-based as well as Arm64 based on AWS Graviton
- Processor types: GPU, CPU, AWS Inferentia and AWS Trainium
- SDK variants: CUDA for NVIDIA GPUs, AWS Neuron for Inferentia and Trainium
- Operating systems: Amazon Linux and Ubuntu
Typical use cases
Model training on GPU instances: Teams launch a GPU instance with a matching DLAMI and start training immediately, without reconciling CUDA versions and framework bindings themselves.
Inference on specialized hardware: DLAMI variants with the AWS Neuron SDK are used for inference on AWS Inferentia and training on AWS Trainium.
Distributed training: The documentation describes scenarios for distributed training across multiple instances.
Interactive experimentation: Through a Jupyter Notebook server on the DLAMI instance, data scientists work directly in the browser.
Benefits
- No manual assembly of driver, library and framework stacks
- Matching variants for CPU, GPU, Inferentia and Trainium
- Available in most AWS Regions
- No additional charge beyond EC2 usage
Integration with innFactory
As an AWS Reseller, innFactory supports you with AWS Deep Learning AMIs: selecting the right AMI variant and instance types, building reproducible training environments, and optimizing cost, for example through EC2 Spot Instances.
Typical Use Cases
Technical Specifications
Frequently Asked Questions
What are AWS Deep Learning AMIs?
AWS Deep Learning AMIs (DLAMI) provide customized machine images for deep learning in the cloud. They are available in most AWS Regions for a variety of Amazon EC2 instance types, from a small CPU-only instance to the latest high-powered multi-GPU instances.
What is preinstalled on a DLAMI?
AWS states that the DLAMIs come preconfigured with NVIDIA CUDA and NVIDIA cuDNN and the latest releases of the most popular deep learning frameworks. That removes the need to install drivers, libraries and frameworks manually.
Which DLAMI variants exist?
AWS groups the images by DLAMI type (Base, Single-Framework, Multi-Framework also known as Conda DLAMI), compute architecture (x86 or Arm64 with AWS Graviton), processor type (GPU, CPU, Inferentia, Trainium), SDK (CUDA or AWS Neuron) and operating system (Amazon Linux or Ubuntu).
How much do the Deep Learning AMIs cost?
There is no separate charge for the AMIs themselves. You pay the standard prices for the Amazon EC2 instances you use and for related AWS resources such as storage and data transfer.
Do the DLAMIs support Inferentia and Trainium?
Yes. AWS provides DLAMI variants with the AWS Neuron SDK designed for AWS Inferentia and AWS Trainium. The documentation includes dedicated tutorials for them.
Note: All product information on this page has been compiled with care, but is provided without guarantee and may be outdated or incomplete. Cloud services evolve rapidly — features, pricing, SLAs, and availability change frequently. Authoritative and up-to-date information can only be found on the official product page of AWS (official documentation). This page does not represent an offer by AWS.