USGS Supercomputers & Research

By Advanced Research Computing (ARC)

High Performance Computing is a rapidly changing field which requires constant research. At any given time ARC is researching multiple new technologies and methodologies in order to create increasingly efficient and user-friendly workflows, and to provide the computing power necessary for the wide array of scientific computing workloads within the USGS.

Computer server racks — Denali is the USGS's flagship supercomputer installed and housed at the EROS Data Center in Sioux Falls, SD. Estimated peak computational performance is expected to be ~660 Tflop/s. Learn more here: USGS Denali Supercomputer

Need more computing power?

We provide USGS scientists with access to high-performance computing (HPC) resources.

There are three on-premises USGS supercomputers housed in Sioux Falls, SD:

Denali is the flagship USGS supercomputer for large-scale models.
Tallgrass is a prototype system designed for artificial intelligence and machine learning workloads.
Hovenweep is our newest supercomputer and replaces Yeti as the workhorse USGS supercomputer for general-purpose HPC workloads.

For more information or to get started using Denali, Tallgrass, or Hovenweep, contact us at hpc@usgs.gov.

Current Research Initiatives

Computer Node — Internal view of a compute node.(Credit: Jeff Falgout. Public domain.)

Accelerator Technologies: Accelerator use has become a standard in modern supercomputing. How can GPUs or other coprocessors be used to benefit scientific research?

Machine Learning: Over the past few years machine learning and deep learning methodologies have taken off and been successful in a variety disciplines, but how can these techniques benefit USGS science?

Emerging Processor Technologies and Software Ecosystems: As new technologies become available, what are the cost savings vs. productivity trade-offs for scientific use cases?

High Performance Data Analytics (HPDA): How can ARC take advantage of the confluence of HPC and Big Data analytics?

Data Management: How can we streamline data management workflows for HPC research? What are the best practices for data movement (transfer) and storage?

Containers: Can containers be used to package entire scientific workflows, software and libraries, and even data? How will they affect reproducibility and portability?