ReleaseBytes
aws AWS What's New ·

AWS ParallelCluster 3.16 Introduces On-Node Diagnostics Tool and Stability Improvements

infraawsgaengineer
feature patch announcement

AWS ParallelCluster 3.16 is now generally available, featuring a new `pcluster-diag` on-node diagnostics tool built into the AMIs for streamlined issue identification. This release also enhances cluster lifecycle resilience, improving the stability of creation, updates, and image builds. Additionally, the HPC and AI/ML software stack has been refreshed with updated NVIDIA driver, CUDA, EFA installer, and Slurm versions. These updates benefit R&D customers and IT administrators operating high-performance computing clusters on AWS.

  • →New on-node diagnostics tool for troubleshooting clusters
  • →Improved cluster lifecycle stability and resilience
  • →Updated HPC and AI/ML software stack components
Features (1) ›
  • New on-node diagnostics tool for troubleshooting clusters

    AWS ParallelCluster 3.16 includes `pcluster-diag`, a diagnostics tool built directly into the ParallelCluster AMIs. It enables users to run diagnostic checks on any cluster node with a single command, generating a structured report to facilitate issue identification.

Enhancements (2) ›
  • Improved cluster lifecycle stability and resilience

    This release hardens the cluster lifecycle by introducing more resilient processes for cluster creation, updates, and image builds, enhancing overall stability and operational reliability.

  • Updated HPC and AI/ML software stack components

    The software stack has been refreshed with newer versions of key components, including the NVIDIA driver, CUDA, EFA installer, and Slurm, providing access to the latest performance and features for HPC and AI/ML workloads.

Read the original announcement →

https://aws.amazon.com/about-aws/whats-new/2026/08/aws-parallelcluster/

Related releases