Case Study: TRIUMF (Canada’s national subatomic physics laboratory) achieves 5X lower storage costs and seamless global ATLAS collaboration with DataDirect Networks SFA 12KXE

Canadian Atlas Tier-1 Data Centre Embeds Dcache To Significantly Speed Global Collaboration

TRIUMF, Canada’s national laboratory and an ATLAS Tier‑1 data centre for the Large Hadron Collider at CERN, supports some 3,000 scientists across 177 institutions in 38 countries and needed high‑performance, petabyte‑scale storage to ingest, process and distribute massive experimental datasets with low latency and high reliability. The challenge was to improve density, performance and global collaboration while driving down space, power and operational costs for continuous, mission‑critical scientific workloads.

TRIUMF deployed DDN SFA 12KXE with embedded dCache and in‑storage processing to consolidate storage and data distribution into a single converged platform, enabling seamless access between Tier‑1 and Tier‑2 sites and support for thousands of concurrent cores. The new architecture cut infrastructure footprint from 21 4U servers to one half‑rack system with eight virtual dCache nodes, reduced power from ~25 kW to under 7 kW, delivered a reported 5× storage infrastructure cost reduction, and added resiliency and encryption for round‑the‑clock scientific collaboration.

TRIUMF

Reda Tafirout

Research Scientist