Research Infrastructure: CC* Data Storage: 20 Petabyte Campus Research Storage Facility at Johns Hopkins University

There is a wide spectrum of ongoing research projects at Johns Hopkins University (JHU) that generate and analyze many petabytes of data. These include both large numerical simulations (turbulence, cosmology, ocean circulation models) and large microscopy data from a broad range of areas (cryoEM, fluorescent microscopy, material science projects, NG sequencing) and various large-scale data fusion projects from digital social science.

The JHU Institute for Data Intensive Engineering and Science has been supporting these activities for the last decade by providing a platform to store and serve petabytes at high speeds and at excellent economies of scale across all disciplines at JHU and beyond. After a workshop held at JHU in the summer of 2022 the need for a new larger-scale, campus-wide research storage facility became clear.

This CC* project builds a 30PB data facility for Big Data projects, in a very cost-effective way, building on previous NSF investments. The system is based on upgrading the 90-node Data-Scope cluster whose Supermicro chassis and power supplies are standardized and compatible with the latest motherboards and CPUs. These nodes are upgraded at a low per node cost to a AMD Ryzen system and turned into a high-performance Ceph cluster with 30PB net storage capacity.

The new storage serves a very wide range of scientific research, including offering many petabytes of open scientific data for global science including SDSS, turbulence, ocean circulation, and cosmology simulations.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.