HPE Cray EX4000 (Hunter)

Photo of HLRS's Hunter supercomputer, with image of Earth weather simulation on the side of the machine.. Copyright: Julian Holzwarth/HLRS

Hunter is HLRS’s newest flagship supercomputer. Based on the HPE Cray EX4000 platform from Hewlett Packard Enterprise, it is a world-class high-performance computing system that enables large-scale, state-of-the-art applications of simulation, artificial intelligence, and data analytics, including hybrid computing approaches that combine diverse methods into powerful workflows.

Hunter marks a milestone for HLRS, as it moves away from the center’s previous CPU-oriented approach to make greater use of GPUs. At its core is the AMD Instinct™ MI300A accelerated processing unit (APU), which combines CPUs, GPUs, and high-bandwidth memory in a single package.

The technology offers fast data transfer speeds, impressive performance, easy programmability, and great energy efficiency. With a theoretical peak performance of 48.1 PFlop/s, Hunter's speed is nearly double that of its predecessor, Hawk, while slashing energy requirements at peak performance by approximately 80%.

HLRS conceived Hunter as a transitional system that will enable system users to prepare for its upcoming supercomputer, Herder. Hewlett Packard Enterprise will also manufacture that future next-generation system, whose installation is scheduled to begin in 2027.

In June 2025, Hunter debuted at number 12 on the Green500 List of the world's most energy efficient supercomputers.

Funding

Funding for the Hunter supercomputer is provided by the German Federal Ministry of Research, Technology and Space and the Baden-Württemberg Ministry for Science, Research and Arts through the Gauss Centre for Supercomputing.

Technical details

Platform

HPE Cray EX4000

Performance

Theoretical peak performance48.1 PFlops

Processor types

APU nodeAMD Instinct MI300A
CPU nodeAMD EPYC 9374F

Number of processors

APUs752 MI300A
CPUs512 AMD EPYC 9374F

Number of nodes

APU188
CPU256

Memory technology

APU512 GB HBM3 (~5.3 TB/s)
CPU768 GB DDR5-4800

Storage

Cray ClusterStor E200025 PB capacity on 2120 disks

Networking

HPE Slingshot 11 Dragonfly(APU: 4 x 200 Gbps per node)

Energy usage (average)

560 kW 

Start of operation

February 2025