Z by HP · Full lineup
WorkstationsBuilt for AI
The Z lineup runs from ultra-compact Z2 Mini to quad-GPU Z8 Fury. 8B to 405B parameters, all air-gap capable. Pick the box that matches your model size.
Pick your tier
Each Z is sized for a different model class. Quantum Pipes runs the same on every one.
405B
Top of the line
Z8 Fury
When you need to fine-tune frontier-scale models on-prem. Quad-GPU tensor parallelism, datacenter-class.
- Quad RTX 6000 Ada
- Up to 1TB RAM
- 60-core Intel Xeon
70B
Power desktop
Z4 G5
The sweet spot for most teams. Runs Llama 3.3 70B and serves a small org without breaking a sweat.
- RTX 6000 Ada
- 256GB RAM
- Threadripper PRO
70B (q4)
Mobile workstation
ZBook Fury
Sovereign AI on a plane. Run quantized 70B models in your lap, completely offline.
- RTX A5000 mobile
- 128GB RAM
- Up to 16TB NVMe
8B
Ultra-compact
Z2 Mini
Pilot-grade. Fits under a monitor. Perfect for testing Quantum Pipes before scaling up.
- RTX A2000
- 64GB RAM
- 1L form factor
Same Quantum Pipes.
Any Z workstation.
Start on a Z2 Mini, scale up to a Z8 Fury when you outgrow it. The stack is identical — only the silicon changes.
Air-gap capable8B – 405B paramsCapsule governance