AWS PCS now provides a production-ready Deep Learning AMI

Amazon Web Services has launched a production-ready Deep Learning AMI (Amazon Machine Image) specifically designed for its Parallel Computing Service (AWS PCS), providing researchers and organizations with a pre-configured foundation for AI/ML training and high-performance computing workloads. The new PCS-ready DLAMI builds on the Deep Learning Base GPU AMI running Ubuntu 24.04 and includes essential infrastructure components such as NVIDIA GPU drivers, CUDA toolkit, EFA drivers, Lustre client, PCS Agent, and Slurm for cluster management. The AMI automatically activates the appropriate Slurm version based on cluster configuration and supports multiple Slurm versions out of the box. AWS maintains and regularly updates the image to include security patches and driver updates, while users can layer additional frameworks and application software on top to customize their environments. The service is available for both x86_64 and arm64 architectures across all AWS regions where PCS is offered, at no additional cost beyond standard compute charges. This release addresses a common pain point in HPC and AI/ML workflows where organizations spend significant time configuring and maintaining cluster infrastructure rather than focusing on research and model development. By providing a managed, pre-tested environment with built-in observability and automatic updates, AWS aims to reduce the operational overhead of running large-scale computational workloads in the cloud.