Overview
AWS Deep Learning AMIs (DLAMI) provides tools to accelerate deep learning in the cloud. The AMIs are preconfigured with popular frameworks, including TensorFlow, PyTorch, and Apache MXNet.
Supported Deep Learning AMIs
For details on the supported AMIs, review the release notes. For containerized AI/ML workloads, see AWS Deep Learning Containers.
| Frameworks: | PyTorch | TensorFlow | |||
| Operating systems: | Ubuntu Linux | Amazon Linux 2 | |||
| Instances: | NVIDIA GPUs | AWS Trainium | AWS Inferentia | ||
| Platforms: | Amazon EC2 | Amazon ECS | Amazon EKS | AWS Graviton |
Accelerate your model training
To expedite your development and model training, DLAMI includes the latest NVIDIA GPU acceleration through preconfigured CUDA and cuDNN drivers, as well as the Intel Math Kernel Library (MKL), popular Python packages, and the Anaconda Platform.
GPU instances - NVIDIA
P3 instances provide up to 14 times better performance than previous-generation Amazon EC2 GPU compute instances. With up to 8 NVIDIA Tesla V100 GPUs, P3 instances provide up to one petaflop of mixed-precision, 125 teraflops of single-precision, and 62 teraflops of double-precision floating point performance.
Powerful compute - Intel
C5 instances are powered by 3.0 GHz Intel Xeon Scalable processors and allow a single core to run up to 3.5 GHz using Intel Turbo Boost Technology. C5 instances offer a higher memory-to-vCPU ratio, deliver 25% improvement in price performance compared to C4 instances, and are ideal for demanding inference applications.
Python packages
DLAMI comes installed with Jupyter Notebook, loaded with Python 2.7 and Python 3.5 kernels and popular Python packages, including the AWS SDK for Python.
Anaconda platform
To simplify package management and deployment, DLAMI installs the Anaconda2 and Anaconda3 Data Science Platform for large-scale data processing, predictive analytics, and scientific computing.
