HP Inc partners with NVIDIA and Red Hat to launch Enterprise AI Platform with 20 PFLOPS performance

HP Inc Unveils Enterprise AI Platform with Red Hat and NVIDIA

HP Inc has announced a strategic collaboration with Red Hat and NVIDIA to develop an open, Enterprise AI Platform designed to run production inference closer to users and data1

1

2

2

. The planned solution brings together HP ZGX Fury, powered by the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip, with Red Hat AI Factory to deliver up to 20 PFLOPS FP4 AI performance for Local AI inference. This collaboration aims to give organizations more choice in where AI workloads run, whether locally, in the cloud, or across Hybrid cloud environments. HP ZGX Fury, based on the NVIDIA DGX Station platform, is now available to order and is certified to run on Red Hat Enterprise Linux.

Addressing Critical Enterprise Requirements

The platform addresses critical enterprise requirements including Latency, Privacy, Data sovereignty, and connectivity needs for organizations deploying AI inference closer to deployment sites1

1

. Jim Nottingham, Senior Vice President and Division President of Advanced Compute and Solutions at HP Inc, emphasized that "The future of AI is moving closer to where people work, machines operate and critical decisions are made." The solution is designed to support multiple AI workloads on the same system while maintaining Workload isolation and operational control. By enabling local processing, organizations can reduce the risks associated with sending sensitive data to external cloud environments while maintaining compliance with regional data regulations.

Technical Architecture and Performance Benefits

Red Hat AI Factory with NVIDIA is built on the infrastructure of Red Hat Enterprise Linux and OpenShift for deploying and managing AI models, agents, and applications across the hybrid cloud2

2

. The platform integrates NVIDIA AI Enterprise with Red Hat AI Enterprise, bringing together co-engineered and jointly validated capabilities. HP's open enterprise-grade AI platform aims to help companies maximize local AI inference throughput, reduce environment setup time and deployment risk, and improve GPU utilization through optimized NVIDIA CUDA libraries, scheduling, and multi-GPU workload orchestration. The solution accelerates AI development by reducing setup time, enabling local agentic coding, and allowing companies to offload compute to the ZGX Fury without altering existing workflows.

Evaluation and Industry Applications

Customers will be able to evaluate the solution in a Sandboxed environment delivered on HP devices running Red Hat AI Factory with NVIDIA before moving use cases into production1

1

2

2

. Details on timing, locations, eligibility, supported configurations, and access will be shared when available. HP ZGX Fury is available via the Red Hat Ecosystem Catalog, ensuring enterprise lifecycle management and support pathways designed to improve consistency from developer environments to production deployment. The collaboration targets industries including manufacturing, healthcare, retail, government, and software development for applications requiring local AI processing capabilities. This matters because organizations in regulated industries can now deploy powerful AI capabilities while maintaining control over their data and meeting compliance requirements.

Implications for Scalable AI Infrastructure

This collaboration signals a shift toward distributed AI deployment models where compute power moves closer to data sources rather than centralizing everything in cloud data centers. The emphasis on Scalable AI infrastructure and AI workload management reflects growing enterprise demand for flexible deployment options. By combining HP's hardware expertise with Red Hat's enterprise software platform and NVIDIA's AI acceleration technology, the partnership creates a comprehensive solution for AI development and deployment. Organizations should watch how this platform performs in real-world production environments and whether it successfully addresses the balance between local processing power and cloud scalability. The 20 PFLOPS performance metric positions this solution competitively for demanding inference workloads, potentially enabling new use cases that were previously impractical due to Latency or data transfer constraints.