Case Study: Durham University achieves scalable, high-performance petabyte HPC storage for cosmological research with Dell Technologies

Durham University works with Dell EMC to build an expansive and scalable HPC storage environment for cosmological research

Durham University needed an efficient, scalable high-performance computing (HPC) storage environment to support cosmological simulations run by its Institute for Computational Cosmology. The university worked with Dell Technologies, deploying solutions such as Ready Solutions for HPC Lustre Storage with ZFS, PowerEdge R640 servers with NVMe, PowerVault ME4024/ME484 arrays, PowerSwitch networking and Mellanox InfiniBand interconnects.

Dell Technologies delivered an end-to-end storage stack for the COSMA 7 cluster that enables very high I/O (up to 200 GB/s) and multi-petabyte capacity (about 2.6 PB usable), while reducing footprint (from ~2.5 racks to two‑thirds of a rack) and cutting power from >22 kW to ~6 kW. The solution accelerated data write/read times for simulations that produce hundreds of terabytes, improving cluster efficiency, lowering CO2 emissions and helping researchers advance cosmological research.

Durham University

Alistair Basden

HPC Technical Manager