NIRD Central Data Library Adds Caching Capabilities

20.05.2026

Following a successful pilot period, we are pleased to announce that the NIRD Central Data Library (CDL) has been enhanced with long-awaited caching capabilities.

Users will be able to pre-cache data on HPC systems before launching jobs that depend on datasets stored in the CDL.

Illustration of a data workflow where datasets from the Central Data Library are pre-cached on HPC systems to enable faster running jobs and improved performance.

The NIRD Central Data Library (CDL) is a complementary service available to users of Sigma2's national research infrastructure, including users of NIRD and HPC resources.

The CDL functions as a centralised repository for research projects, supporting the storage of shared data assets across multiple projects, including input datasets, software libraries, AI models, and AI training data.

Full details on the procedure

Key Improvements

  • Central, master copy on NIRD

  • Improved data management

  • Smaller footprint on HPC machines