Home » Best practices

Best practices

Best Practices

To support continuous improvement and shared learning, the EPICURE portal features a curated selection of best practices materials. These resources are designed to promote consistency, efficiency, and quality, while encouraging innovation and collaboration.

Best practices: Getting your code working in the Large Scale

Optimize large-scale HPC workloads with CPU affinity, MPI/UCX tuning, profiling, and debugging.

Best Practice Guide on Artificial Intelligence on EuroHPC Systems featured image

Best Practice Guide on Artificial Intelligence on EuroHPC Systems

Guide to deploying and scaling AI workloads on EuroHPC systems, with tools, benchmarks, scripts, and best practices.

Benchmarking and using AlphaFold 3 on Leonardo

AF3 on Leonardo: benchmarks, HMMER scaling, and Slurm tuning for efficient HPC workflows.

Best Practice Guide to Scaling PyTorch DistributedDataParallel (DDP) Training to Multiple Nodes on the Vega Supercomputer

A best practice guide to scaling PyTorch DDP training to multiple nodes on HPC systems on Vega.

Best Practice Guide on Power Consumption Measurements in EuroHPC Systems featured image

Best Practice Guide on Power Consumption Measurements in EuroHPC Systems

An overview of how users can access and interpret power consumption data across currently active EuroHPC supercomputers.

Best Practice Guide to Running AI Workloads on Leonardo CINECA

A guide on how to set up and run AI workloads on the Leonardo supercomputer without using containers.