HPE and NVIDIA unveiled an expanded portfolio of enterprise AI infrastructure at GTC 2025, including new AI-optimized servers, private cloud solutions, modular data center designs, and AI-native observability tools. The expanded NVIDIA AI Computing by HPE platform is designed to help enterprises accelerate AI model training, fine-tuning, and inferencing, while improving performance, security, and power efficiency for both on-premises and hybrid deployments.
HPE introduced enhancements to its Private Cloud AI, now integrated with the NVIDIA AI Data Platform. The platform provides enterprises with a turnkey, self-service cloud experience for deploying generative, agentic, and physical AI workloads. The Private Cloud AI portfolio now includes a new developer system, unified edge-to-cloud data management via HPE Data Fabric Software, and support for rapid deployment of pre-validated NVIDIA blueprints, such as the Multimodal PDF Data Extraction Blueprint and Digital Twins Blueprint.
In parallel, HPE expanded its AI server lineup with several new ProLiant Compute models built around the latest NVIDIA Blackwell Ultra and Grace Blackwell platforms. These include the GB300 NVL72 by HPE for massive AI clusters, the ProLiant Compute XD with HGX B300, and the ProLiant DL384b Gen12 with GB200 NVL4 Superchips. HPE also introduced the AI Mod POD, a modular, liquid-cooled, 1.5MW-capable data center for accelerated workloads, and enhanced its OpsRamp platform with GPU observability and optimization tools for AI-native workloads running across NVIDIA clusters.
• HPE expands Private Cloud AI to integrate with the NVIDIA AI Data Platform for hybrid AI deployments.
• New developer system adds AI-ready environments with 32TB storage and NVIDIA accelerated computing.
• Launches new AI-optimized servers with NVIDIA Blackwell Ultra, GB300 NVL72, and Grace Blackwell NVL4 technologies.
• Introduces HPE AI Mod POD: a modular 1.5MW-capable liquid-cooled data center for AI and HPC workloads.
• HPE OpsRamp now includes GPU observability and optimization for NVIDIA clusters.
“AI is delivering significant opportunity for enterprises, and requires a portfolio of streamlined and integrated solutions to support widespread adoption,” said Antonio Neri, president and CEO at HPE.
