Summary:
- Hewlett Packard Enterprise is expanding its HPE Cray supercomputing lineup with new blades, storage, interconnect, and management software to meet the increasing demands of AI and high-performance computing.
- The updated platform offers a unified architecture for research labs, sovereign computing initiatives, and large enterprises looking to integrate AI and simulation workloads.
- European research centers like HLRS and LRZ have chosen the HPE Cray GX5000 platform for their flagship systems, emphasizing performance, energy efficiency, and sustainability.
Hewlett Packard Enterprise (HPE) is enhancing its HPE Cray supercomputing range with a range of new components to address the growing computational and energy needs of large-scale AI and high-performance computing (HPC) tasks. The updated platform is designed to cater to research labs, sovereign computing projects, and large businesses that seek to merge AI and simulation workloads on a single infrastructure rather than segregating them.
The latest additions to the HPE Cray lineup complement the introduction of the HPE Cray Supercomputing GX5000 platform and the K3000 storage system. These enhancements aim to deliver increased compute density, improved energy efficiency, and enhanced operational control to manage the growing size and complexity of AI models. Organizations are now focusing on managing power consumption, integrating AI with existing workflows, and ensuring infrastructure security in multi-tenant environments.
Leading European research centers like HLRS and LRZ have already selected the updated platform for their flagship systems. HLRS expects a significant performance boost for simulation and AI workloads while reducing the energy footprint of its data center. On the other hand, LRZ is prioritizing sustainability with its Blue Lion system, featuring 100% direct liquid cooling and designed for higher performance and tighter integration of modeling, simulation, and AI.
The core of the portfolio expansion includes three new directly liquid-cooled compute blades supporting a variety of CPUs and GPUs from different vendors. Each blade type caters to specific workloads while maintaining compatibility within the same chassis and management framework. The HPE Cray Supercomputing GX440n Accelerated Blade targets NVIDIA platforms for AI and HPC, the GX350a Accelerated Blade is designed for AMD ecosystems, and the GX250 Compute Blade suits CPU-only workloads requiring high double-precision performance.
In addition to hardware upgrades, HPE is launching new Supercomputing Management Software to facilitate the management of complex systems, supporting multi-tenant environments, virtualization, and containerization. This software enables operators to host various user communities and workload types while ensuring isolation where necessary. Energy awareness, enhanced security controls, and governance reporting are also integrated to align with sovereign computing projects and regulatory requirements.
Interconnect performance is a key focus, with the introduction of the Slingshot 400 interconnect optimized for dense, liquid-cooled form factors and large AI/HPC installations. The HPE Cray Supercomputing Storage Systems K3000, based on HPE ProLiant DL360 Gen12 servers, integrates the Distributed Asynchronous Object Storage (DAOS) stack for low latency and high throughput, ideal for AI training pipelines and data-intensive simulations.
HPE also offers a range of services around the Cray line, including application performance tuning, deployment, and operational support. Partners like AMD and NVIDIA highlight the collaborative efforts with HPE to deliver scalable, energy-efficient systems for scientific and AI workloads. The staggered availability of the new components indicates HPE’s alignment with future AI and exascale-class procurements, allowing early adopters to plan for converged architectures that blend AI, simulation, and data workflows seamlessly.
