New supercomputing portfolio features direct liquid-cooled multi-partner, multi-workload blades, unified management software and modern, high-performance interconnect to help customers achieve productivity at-scale in the era of converged AI and HPC workloads

In this article

• HPE introduces three multi-partner, multi-workload compute blades for its next-generation supercomputing platform achieving one of the highest compute performance densities in the industry
• The HPE Supercomputing Management Software delivers a secure systems management experience for AI and HPC infrastructure across all phases of the system life cycle
• University of Stuttgart and Leibniz Supercomputing Centre select the new HPE Cray Supercomputing GX5000 platform, validating industry leadership innovation in the new solutions

HPE (NYSE: HPE) today announces the newest additions to the next-generation HPE Cray supercomputing portfolio that provide industry-leading compute density¹, designed to meet the needs of artificial intelligence (AI) demands while performing at-scale. This HPE Cray Supercomputing portfolio expansion includes three multi-partner, multi-workload compute blades, unified management software and high-performance interconnect support, establishing one of the most powerful supercomputing architectures in the industry for research laboratories, sovereign entities and large enterprises that are increasingly turning to AI to advance their simulations and drive scientific discovery.

“Global organizations relying on supercomputing are looking for better computing performance for all of their workloads,” said Trish Damkroger, senior vice president and general manager, HPC and AI Infrastructure Solutions at HPE. “Our new HPE Cray Supercomputing platform is the answer to customers’ needs for higher performance density with a unified AI and HPC architecture that is engineered for groundbreaking outcomes. HPE is proud to share our unwavering commitment to AI and supercomputing as the engine of innovation and scientific discovery, advancing the way people live and work.”

This portfolio expansion is a follow up to last month’s debut of HPE Cray Supercomputing GX5000 that is purpose-built to enable supercomputing breakthroughs in the converged AI and HPC era. The platform is augmented by the new HPE Cray Supercomputing Storage Systems K3000, which is the industry’s first factory-built storage system with embedded Distributed Asynchronous Object Storage (DAOS) open source software for increased performance².

Next-Gen HPE Cray Platform Sees Rapid Industry Adoption

The High-Performance Computing Center of the University of Stuttgart (HLRS) and Leibniz Supercomputing Centre (LRZ) of the Bavarian Academy of Sciences and Humanities have selected the HPE Cray Supercomputing GX5000 as the platform for their next generation supercomputers. This next-generation supercomputing portfolio is already gaining traction among some of the world’s highest performing supercomputers, including the upcoming Herder supercomputer from HLRS and Blue Lion supercomputer from LRZ, enabling each customer to unlock new scientific horizons.

"As the builder of HLRS's Hawk and Hunter supercomputers, HPE has for years been an excellent partner for HLRS,” said Prof. Dr.-Ing. Michael Resch, director at HLRS. “HPE's supercomputer systems and expert support have directly enhanced the ability of our scientific and industrial user communities to make novel scientific discoveries and design better technologies. As we look forward to the arrival of Herder, we are excited to begin the next stage of this cooperation with HPE. The GX5000 platform will offer our users a major leap in performance for simulation and artificial intelligence, as well as improved energy efficiency — a primary concern at our HPC center."

“At LRZ, our mission is to deliver world-class high-performance computing that drives research with global impact, while ensuring our supercomputers operate efficiently and sustainably,” said Prof. Dr. Dieter Kranzlmüller, chairman of the board of directors at LRZ. “The upcoming HPE Cray GX500 platform, chosen for our next flagship system Blue Lion, perfectly reflects this commitment. Featuring 100 percent direct liquid cooling capable of operating at temperatures of up to 40 °C, the system enables the reuse of waste heat across the Garching research campus. With sustained performance up to 30 times faster than our current system, Blue Lion will empower researchers to seamlessly integrate traditional modeling and simulation workflows with cutting-edge AI methods—pushing the boundaries of scientific discovery like never before.”

New direct liquid-cooled multi-partner, multi-workload processing blades

Supported by 100 percent direct liquid cooling, these three new blades allow customers to deploy industry-leading performance density for flagship GPUs like the next-generation NVIDIA Rubin platform and AMD Instinct™ MI430X or flagship CPUs like the next-gen AMD EPYC™ processors, codenamed “Venice”. All processing blades feature a choice of either four or eight HPE Slingshot 400 gigabits per second (Gbps) endpoints per blade and an option for two non-volatile memory express (NVMe) solid-state drives (SSDs) per blade. A mix of the three processing blades can deploy in a HPE Cray Supercomputing GX5000 compute rack to enable customers to configure the optimal mix of blades for their workload, even in small configurations.

• HPE Cray Supercomputing GX440n Accelerated Blade – This blade offers a universal compute engine for mixed-precision computing with four NVIDIA Vera CPUs and eight NVIDIA Rubin GPUs. Up to 24 of these blades can be configured per compute rack, delivering industry-leading NVIDIA Rubin GPU density with up to 192 NVIDIA Rubin GPUs per rack.
• HPE Cray Supercomputing GX350a Accelerated Blade – This blade is recommended for customers who desire a universal compute engine for mixed-precision computing with AMD CPUs and GPUs. The blade includes one next-generation AMD EPYC, “Venice”, CPU and four AMD Instinct™ MI430X GPUs, which is the new MI400 series accelerator engineered specifically for sovereign AI and HPC. Up to 28 of these blades can be configured per compute rack, providing up to 112 AMD MI430X GPUs per rack.
• HPE Cray Supercomputing GX250 Compute Blade – Designed for next generation supercomputers, this blade is recommended for customers who desire to build a CPU-only partition for double-precision workloads with eight next-generation AMD EPYC, “Venice”, CPUs. In this case, the GPU partition of the next generation supercomputer could be built on either one of the previously listed accelerated blades – depending on the preference of the customer. Up to 40 of those blades can be configured per compute rack, delivering industry-leading flagship x86 CPU core density.

A unified, multi-tenant and secure systems management experience

The HPE Supercomputing Management Software introduces new capabilities to support multi-tenant, virtualized and containerized environments to give customers enhanced flexibility to deploy converged AI and HPC processing, isolating workloads and user groups when necessary. The new software also provides management of system-wide power and energy. This feature allows customers to monitor power usage, help estimate consumption, and integrate with power-aware schedulers, enabling customers to maximize energy efficiency and anticipate costs.

The software delivers a unified and secure systems management experience for AI and HPC infrastructure across all phases of the supercomputer’s life cycle, including provisioning, monitoring, power and cooling, and scaling. The new capabilities come with strengthened security features and improved governance reporting.

HPE Slingshot 400 now available for GX5000-based systems

Now available for the HPE Cray Supercomputing GX5000 platform, HPE Slingshot 400 has been engineered for the denser form factor and is designed to perform at-scale under the heavier workloads that the new converged AI and HPC platform enables.

A new chassis was developed for the 100 percent direct liquid-cooled switch blade, which features 64 400 Gbps ports, and is deployed in the following configurations:

• 8 switches with 512 ports
• 16 switches with 1,024 ports
• 32 switches with 2,048 ports

First announced last year, the HPE Slingshot 400 switches utilize all the available bandwidth deployed in the higher performance topology of the HPE Cray Supercomputing GX5000 and give customers reduced latency, improved sustained bandwidth, and reliability while managing costs.

DAOS-based storage solution ramps up productivity for AI applications

The HPE Cray Supercomputing Storage Systems K3000 is based on the HPE ProLiant Compute DL360 Gen12 server. This storage system delivers industry-leading compute performance, memory density with scalability and high-speed data transfer rates to run demanding applications. These capabilities in combination with the low-latency architecture of DAOS will enable supercomputing customers to run input/output-bound AI applications with higher productivity.

Depending on project requirements, the DAOS storage servers based on HPE ProLiant Compute servers will be available in multiple drive sizes and configurations, including:

• Performance-optimized DAOS storage server with either eight, 12 or 16 non-volatile memory express (NVMe) drives
• Capacity-optimized DAOS storage server with 20 NVMe drives
• Drive sizes of 3.84 TB, 7.68 TB or 15.36 TB
• 512 GB, 1,024 GB or 2,048 GB DRAM configurations based on selected drive size

Connectivity options are HPE Slingshot 200, HPE Slingshot 400, InfiniBand NDR or 400 Gigabit per second Ethernet.

Service experience designed for supercomputing

From application performance optimization to global turnkey implementation and 24x7 operational support, HPE supercomputing services help customers gain and maintain the full value of their investment.

As a global leader in supercomputing³, HPE draws upon decades of expertise to deliver advanced and energy-efficient AI and HPC solutions. By providing unparalleled performance and scalability, HPE empowers customers with the resources they need to realize their most ambitious goals and drive innovation forward.

Availability

• The HPE Cray Supercomputing GX440n Accelerated Blade, HPE Cray Supercomputing GX250 Compute Blade and HPE Cray Supercomputing GX350a Accelerated Blade will be available in early 2027.
• The HPE Supercomputing Management Software will be available in early 2027.
• HPE Slingshot 400 will be available for clusters based on HPE Cray Supercomputing GX5000 systems in early 2027.
• HPE Cray Supercomputing Storage Systems K3000 with HPE ProLiant Compute servers will be available in early 2026.

Supporting quotes

“The latest HPE Cray supercomputing platform reflects the deep collaboration between AMD and HPE to develop leadership technologies at the convergence of HPC and AI,” said Travis Karr, corporate vice president, HPC and Sovereign AI at AMD. “By bringing together high-density compute blades powered by AMD EPYC CPUs and AMD Instinct GPUs, we’re helping customers tackle some of the world’s most complex scientific and AI challenges with scalable, energy-efficient solutions.”

“HPE’s next-generation AI supercomputers with NVIDIA Vera Rubin are designed to supercharge scientific discovery,” said Dion Harris, senior director, HPC, Cloud and AI Infrastructure at NVIDIA. “Powered by NVIDIA, these HPE GX5000 systems will boost simulation, analytics and AI, providing essential infrastructure to accelerate the AI industrial revolution.”

“High performance computing and AI have become very high growth sectors," said Dr. Earl Joseph, CEO at Hyperion Research. "HPE’s next generation of GX5000 supercomputers will help researchers and companies dramatically develop better products and new scientific discoveries, as well as help transform society by addressing critical societal goals.”

About HPE

HPE (NYSE: HPE) is a leader in essential enterprise technology, bringing together the power of AI, cloud, and networking to help organizations achieve more. As pioneers of possibility, our innovation and expertise advance the way people live and work. We empower our customers across industries to optimize operational performance, transform data into foresight, and maximize their impact. Unlock your boldest ambitions with HPE. Discover more at www.hpe.com.

¹ HPE internal analysis of GPU and CPU performance density per rack with the densest, liquid cooled offerings of the next three competitors in the HPC/AI market, September 2025

² DAOS-based storage systems are ranked #1 and #2 on the global IO500 storage benchmark and together have four times the storage benchmark score than the next 30 storage systems

³ Hyperion Research Q4 2023 HPC Market Data Report reflecting CY2023, Supercomputer Segment (May 29, 2024), Hyperion Research

View source version on businesswire.com: https://www.businesswire.com/news/home/20251113871389/en/