HPE Cray Supercomputing
Exascale-class power for your biggest HPC and AI breakthroughs
Overview
HPE Cray Supercomputing brings exascale-class performance to the most demanding HPC and AI workloads. It pairs dense CPU and GPU compute, high-throughput storage, and a complete software stack in one architecture you can run across a flagship supercomputer or distributed clusters. The same systems power more than half of the world's 100 fastest supercomputers, so you get research-grade speed with a clear path to scale.
Why HPE Cray Supercomputing
Converged HPC and AI in one architecture
Run modeling, simulation, and AI training on the same platform, with the highest CPU and GPU density per compute rack and top I/O performance per storage rack.
Proven at the top of the rankings
The HPE Cray EX4000 powers more than half of the world's top 100 most powerful supercomputers, so you are building on hardware trusted by leading labs.
Built to scale beyond exascale
The next-generation GX5000 is already chosen to succeed a verified exascale system, giving you headroom as workloads and data sets keep growing.
Storage tuned for AI and HPC
K3000 factory-built storage embeds DAOS and plugs into PyTorch and TensorFlow, while E2000 scales on Lustre with no per-drive or per-capacity software licensing.
Efficient by design
Direct liquid cooling and high performance density deliver strong price/performance and lower energy use as you grow your footprint.
A complete developer toolchain
The HPE Cray Programming Environment gives your teams tools to develop, port, debug, and tune code so applications reach production faster.
At a glance
- Flagship system
- HPE Cray Supercomputing GX5000
- Current-gen system
- HPE Cray Supercomputing EX4000
- Storage systems
- K3000 (DAOS) and E2000 (Lustre)
- Cooling
- Direct liquid cooling
- AI framework support
- PyTorch, TensorFlow
- Programming environment
- HPE Cray Programming Environment
- Workload focus
- Converged HPC and AI
Where it fits
Frequently asked
What workloads is HPE Cray Supercomputing built for?
It targets the most demanding HPC and AI workloads, including large-scale simulation, modeling, and AI training. The converged architecture lets you run both on the same systems.
How does it compare to other supercomputers?
HPE Cray systems power more than half of the world's top 100 most powerful supercomputers, per the June 2025 TOP500 list. The lineup includes both the current EX4000 and the next-generation GX5000.
What storage options are available?
The K3000 is a factory-built system that embeds DAOS software and integrates with AI frameworks like PyTorch and TensorFlow. The E2000 uses the Lustre parallel file system and scales without per-drive or per-capacity software licensing.
Does it support AI development tools?
Yes. The HPE Cray Programming Environment provides a full set of tools for developing, porting, debugging, and tuning code, and the storage stack integrates directly with common AI frameworks.
Can the same architecture run across different system sizes?
Yes. You can deploy the same unified HPC/AI architecture site-wide, from a flagship supercomputer to distributed HPC and AI clusters.
Related in Supercomputing & AI
Build your HPE bill of materials.
Send us the requirement, the project, or an existing quote to beat. We come back with a validated, TAA-compliant HPE configuration and a real price, often below list.
connect@getuniqcli.com · Chicago, IL