Product Description
HP ZGX Fury AI Station – Enterprise AI Development & Inference Platform
Accelerate AI development, inference, and agentic AI workflows with the HP ZGX Fury AI Station, a powerful on-prem AI workstation designed for enterprise-scale deployment. Powered by the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip, the ZGX Fury enables organizations to run large AI models locally while maintaining data security, predictable operating costs, and complete infrastructure control.
Ideal for AI teams, developers, data scientists, and enterprise organizations, the HP ZGX Fury supports production AI workloads and concurrent users with frontier-class AI models.
Key Benefits
✅ Priority Access Program
Reserve early access before general availability and secure your place in the first production allocations.
✅ Enterprise-Scale AI Performance
Run AI development, production inference, retrieval-augmented generation (RAG), and agentic AI applications on-premises.
✅ Massive Unified Memory
Featuring up to 748GB of coherent memory, including:
- Up to 496GB LPDDR5X CPU memory
- Up to 252GB HBM3e GPU memory
- Up to 7.1TB/s memory bandwidth
✅ NVIDIA GB300 Grace Blackwell Ultra Architecture
Built on the latest NVIDIA AI computing platform for advanced large language model development and deployment.
✅ Secure On-Prem AI Infrastructure
Keep sensitive data local while reducing reliance on cloud resources.
✅ AI-Ready Software Stack
- Ubuntu Operating System
- NVIDIA AI Developer Tools
- NVIDIA AI Software Stack
- HP ZGX Toolkit
How the Priority Access Program Works
Step 1: Reserve Your Spot
Sign up for updates or place a refundable reservation.
Step 2: Receive Product Updates
Stay informed about launch timelines, product availability, and configuration options.
Step 3: Confirm Your Configuration
Select your preferred build and complete your purchase when the platform launches.
Step 4: Deploy AI On-Prem
Bring enterprise AI workloads in-house with high-performance local computing.
Technical Specifications
Best For
- AI Development
- Production AI Inference
- Agentic AI Workflows
- Large Language Models (LLMs)
- Multi-User AI Environments
- Enterprise AI Deployment
Operating System
- Ubuntu with NVIDIA AI Developer Tools
Processor
- NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip
Memory
- 748GB Coherent Memory
- CPU: Up to 496GB LPDDR5X (396 GB/s)
- GPU: Up to 252GB HBM3e (7.1 TB/s)
Included Software
- HP ZGX Toolkit
- NVIDIA AI Software Stack