Case Study: North German Supercomputing Alliance (HLRN) achieves peta-scale performance and accelerated scientific breakthroughs with DataDirect Networks high-performance storage

HLRN - Customer Case Study

The North German Supercomputing Alliance (HLRN) provides peta-scale compute and storage resources to researchers across seven North German states, supporting more than 300 scientists on roughly 80 simultaneous projects. HLRN needed to balance massive compute demand with high‑bandwidth, scalable storage that could be accessed reliably from geographically distributed sites (Berlin and Hannover), provide policy-driven snapshots, and grow from petabytes to tens of petabytes without becoming an I/O bottleneck.

HLRN partnered with Cray and DataDirect Networks to deploy a distributed system based on Cray XC30 and DDN SFA12K-40 with Lustre and DDN GRIDScaler (GPFS) mirrored between sites. The solution delivered nearly four times the previous bandwidth (aggregate >50 GB/s), immediate capacity (4 PB at RRZN, 1.6 PB migrated at ZIB) with plans to scale to 10+ PB, robust snapshot and DirectProtect reliability, lower energy/TCO, and faster time‑to‑science across a broader set of applications.

HLRN

Steffen Schulze-Kremer

Head of HPC department