HPE
HPE brings the Cray lineage to enterprise AI: genuine in-house direct-liquid-cooling engineering proven on exascale systems, applied to ProLiant and Cray XD NVIDIA platforms up to GB300 NVL72 scale. Strong where the buyer wants supercomputing discipline with OEM support.
- Cray supercomputing heritage (acquired 2019): the source of HPE's HPC and DLC engineering stack
- Industry's first 100% fanless direct liquid cooling architecture (Oct 2024): 8-element design covering GPU, CPU, blade, storage, fabric, rack, pod and CDU; claims 90% cooling-power reduction vs air
- Slingshot: proprietary Cray-developed interconnect used in DOE exascale systems
- iLO BMC: HPE designs its own iLO ASIC with silicon root of trust
- Own motherboard and component manufacturing at its Aguadilla, Puerto Rico plant
| Model | Platform | Form factor | Cooling |
|---|---|---|---|
| HPE Cray XD670 8x SXM5 GPUs; InfiniBand, Ethernet or Slingshot fabric | HGX H100/H200 | 5U | air or DLC |
| NVIDIA GB200 NVL72 by HPE 13.4TB unified GPU memory, ~132kW nominal | GB200 NVL72 | rack-scale | DLC |
| NVIDIA GB300 NVL72 by HPE Blackwell Ultra, available 2H 2025 | GB300 NVL72 | rack-scale | DLC |
| ProLiant Compute DL380a Gen12 H200 NVL, L40S, RTX PRO 6000 Blackwell; enterprise entry point | PCIe | 4U | air |
| NVIDIA Vera Rubin NVL72 by HPE Announced for December 2026; Compute XD700 slated early 2027 | Vera Rubin NVL72 | rack-scale | DLC |
- NVIDIA AI Computing by HPE: joint portfolio since HPE Discover 2024
- System builder for GB200 NVL72 (shipped) and GB300 NVL72, plus HGX B200/B300 in ProLiant Compute XD
- Named in NVIDIA's Vera Rubin builder lists; Vera Rubin NVL72 by HPE announced for December 2026
- ProLiant Compute DL380a Gen12 is NVIDIA-Certified incl. NVIDIA Confidential Computing
- Aguadilla, Puerto Rico (HPE-owned plant building motherboards and components)
- Spring, Texas (HQ campus with product development and labs)
- $6.8B new AI systems orders in FY2025, over 60% of cumulative AI orders from sovereign and enterprise customers rather than hyperscalers
- Built El Capitan at LLNL, world's fastest supercomputer from Nov 2024 (1.742 exaFLOPS, 100% fanless DLC)
- GPU-optimized ProLiant configurations quoted at 32-52 week lead times in 2025 (secondary source)
- Differentiates on enterprise relationships, global services and Cray-derived DLC
Is HPE an OEM, ODM or integrator for NVIDIA AI servers?
HPE sits in the "Global OEMs" tier of our index. Cray supercomputing heritage (acquired 2019): the source of HPE's HPC and DLC engineering stack
Does HPE design its NVIDIA servers in-house?
HPE brings the Cray lineage to enterprise AI: genuine in-house direct-liquid-cooling engineering proven on exascale systems, applied to ProLiant and Cray XD NVIDIA platforms up to GB300 NVL72 scale. Strong where the buyer wants supercomputing discipline with OEM support.
Which NVIDIA platforms does HPE ship?
HPE ships systems on HGX H100/H200, GB200 NVL72, GB300 NVL72, PCIe, Vera Rubin NVL72. The systems table on this page lists the specific models.
Where does HPE manufacture its AI servers?
Aguadilla, Puerto Rico (HPE-owned plant building motherboards and components). Spring, Texas (HQ campus with product development and labs)