Kestrel Computing System

The Kestrel high-performance computing (HPC) system has expanded NREL's supercomputing capacity to 44 petaflops.

System Architecture

The Kestrel system is an HPC system for computing intensive and parallel jobs. User facing login and data analysis and visualization nodes will utilize Red Hat Enterprise Linux. Compute nodes will also utilize Red Hat Enterprise Linux. The system utilizes a high-speed, 200 gigabit (Gb)/s Hewlett Packard Enterprise Slingshot interconnect.

Learn about the Kestrel cluster (GitHub).

System Configuration

To learn about Kestrel's hardware, storage, and networking, see Kestrel system configuration.

Running on the System

Learn about Kestrel job partitions and scheduling policies (GitHub).

Filesystems

See the Kestrel filesystems (GitHub).

Programming Environments

See Kestrel programming environments overview (GitHub).

Countdown to Kestrel

Read blogs about the installation of NREL's Kestrel HPC system.