Supercomputing and storage

Computational systems | Data storage resources

CISL provides computing, data storage, data analysis, visualization, networking and other systems customized to support the atmospheric and related sciences.

The computational systems and data storage resources that CISL manages serve a number of purposes. They enable leading-edge research in Earth system science as well as research for the advancement of computer science and high-performance computing.

User documentation is provided on the NCAR Advanced Research Computing (ARC) portal. Links are included below to ARC pages for some individual systems.

Computational systems

Derecho 

The Derecho supercomputer is a 19.87-petaflops system that is expected to deliver about 3.5 times the scientific throughput of the Cheyenne system when it becomes operational in 2022. Derecho will get 20% of its sustained computing capability from graphics processing units (GPUs), with the remainder coming from traditional multicore processors. You can learn more about the system at these links:

Derecho logo on a background designed for the system's "skin."

A rendering of the skin design for the Derecho supercomputer, which will become operational in 2022.

Cheyenne

The Cheyenne supercomputer at the NCAR-Wyoming Supercomputing Center.

The Cheyenne supercomputer at the NCAR-Wyoming Supercomputing Center.

Cheyenne is a 5.34-petaflops HPC system that has 145,152 processor cores on 4,032 nodes. This supercomputer was installed at the NCAR-Wyoming Supercomputing Center in late 2016 and became operational in January 2017.

User documentation.

Casper

The Casper cluster is a heterogeneous system of specialized data analysis, visualization resources and large-memory, multi-GPU nodes. In addition to its analysis and visualization capabilities, Casper supports explorations in machine- and deep-learning in atmospheric and related sciences.

User documentation.

Data storage resources

GLADE: Centralized file systems and data storage

GLADE consists of file system servers and storage devices that are shared by the Cheyenne supercomputer and the Casper nodes. The centralized file systems allow scientists to generate model output on the supercomputer, then analyze or visualize it on data analysis and visualization nodes, eliminating the need to move data between clusters.

User documentation.

Campaign Storage

The NCAR Campaign Storage file system is a disk-based resource for medium-term storage of project data by NCAR labs and universities.

User documentation

Quasar archive

The Quasar Archive is a cold archive for storing curated data collections that have an indefinite lifetime. It is not designed to serve data or to store data that will be frequently accessed, overwritten or deleted. 

User documentation

Stratus object storage

Stratus, the CISL object storage disk system, is for long-term data storage.

User documentation.