Case Study: HLRN achieves 16 PetaFLOP/s HPC performance with Cornelis Networks Omni-Path and Intel Xeon processors

HLRN chooses 2nd Generation Intel® Xeon® Scalable processors to meet their increasingly diverse needs for high performance computing workloads

HLRN, a supercomputing alliance supporting over 100 universities and research institutions, faced the challenge of procuring a new supercomputer to meet their increasingly diverse and compute-intensive scientific workloads. They required a system that delivered the highest real-world application performance, not just theoretical peak performance, with the right combination of processor and interconnect being crucial. Cornelis Networks, along with its technology, was selected to meet this demand.

The solution implemented by Cornelis Networks and its partners featured a massive system powered by 2nd Gen Intel Xeon Scalable processors and interconnected with Cornelis Networks' Omni-Path Architecture. This homogeneous system provided the flexibility and high performance needed for HLRN's diverse research. The result was the 16 PetaFLOP/s HLRN-IV system, which is approximately six times faster than its predecessor, enabling groundbreaking research in climate science, fluid dynamics, and healthcare without requiring researchers to change their existing code.

HLRN

Alexander Reinefeld

Professor