Enterprise multi-GPU training and large model inference
HP Z8 Fury G5 (4x RTX 6000 Ada)
HP · Verified 2026-07-23
From $45,000
192GB combined VRAM over PCIe P2P (no NVLink on RTX 6000 Ada). Enterprise training and 405B-class inference with full warranty.
Main limitation: Very expensive ($45k+); large physical footprint; high power requirements
Some links are affiliate links. We may earn a commission at no extra cost to you. Full disclosure.
Key Specifications
| Specification | Value |
|---|---|
| Form Factor | deskside workstation |
| GPU | NVIDIA RTX 6000 Ada Generation |
| GPU Count | 4 |
| GPU Memory per GPU | 48 |
| Total Accelerator Memory | 192 |
| Memory Architecture | dedicated VRAM (GDDR6) |
| GPU Memory Type | GDDR6 |
| Memory Bandwidth (per GPU) | 960 |
| System RAM | 256 |
| RAM Type | DDR5-4800 ECC |
| ECC Memory | Yes |
| Max System RAM | 2048 |
| CPU | Intel Xeon w9-3495X |
| CPU Cores | 56 |
| CPU TDP | 350 |
| Primary Storage | 8192 |
| Storage Type | NVMe SSD RAID |
| PCIe Generation | PCIe 5.0 |
| PCIe x16 Slots | 4+ |
| NVLink Support | No (RTX 6000 Ada has no NVLink — NVIDIA datasheet 2504660) |
| Power Supply | 1450 (dual) |
| Est. Peak Power | 2400 |
| Cooling | professional air cooling (Z Cool) |
| Dimensions | Full-tower workstation |
| Weight | 40+ |
| OS | Windows 11 Pro / Ubuntu |
| Linux Support | Yes (certified) |
| Windows Support | Yes |
| WSL Support | Yes |
| CUDA Support | Yes |
| Metal Support | No |
| Warranty | 3 |
| Onsite Support | Yes (HP Care Pack) |
| Release Year | 2024 |
What Models Can It Run?
Calculated estimates using weight-only memory analysis at Q4 (4-bit) quantization. Actual requirements vary by framework, context length, and runtime overhead. See methodology.
| Model | Params | Required Memory (Q4) | Fits in 192GB? |
|---|---|---|---|
| Llama 3.1 8B | 8B | 5.52 GB | ✓ Yes |
| Llama 3.1 70B | 70B | 37.14 GB | ✓ Yes |
| Llama 3.2 1B | 1B | 2 GB | ✓ Yes |
| Llama 3.2 3B | 3B | 3.01 GB | ✓ Yes |
| Qwen 2.5 7B | 7B | 5.01 GB | ✓ Yes |
| Qwen 2.5 14B | 14B | 8.53 GB | ✓ Yes |
| Qwen 2.5 32B | 32B | 18.06 GB | ✓ Yes |
| Qwen 2.5 72B | 72B | 38.14 GB | ✓ Yes |
| Mistral 7B | 7B | 5.01 GB | ✓ Yes |
| Mixtral 8x7B | 46.7B | 25.44 GB | ✓ Yes |
| DeepSeek R1 7B | 7B | 5.01 GB | ✓ Yes |
| DeepSeek R1 32B | 32B | 18.06 GB | ✓ Yes |
| DeepSeek R1 70B | 70B | 37.14 GB | ✓ Yes |
| Phi-3 Medium 14B | 14B | 8.53 GB | ✓ Yes |
| Gemma 2 9B | 9B | 6.02 GB | ✓ Yes |
| Gemma 2 27B | 27B | 15.05 GB | ✓ Yes |
| Stable Diffusion XL | 6.6B | 4.81 GB | ✓ Yes |
| Flux.1 Dev | 12B | 7.52 GB | ✓ Yes |
| Flux.1 Schnell | 12B | 7.52 GB | ✓ Yes |
Max model size: 70B class models (~377B params at Q4). These are calculated estimates, not measured results.
Pros & Cons
Pros
- ✓ Very large accelerator memory (192GB)
- ✓ Linux support
- ✓ Windows support
- ✓ ECC memory for reliability
- ✓ Onsite support available
- ✓ 3-year warranty
Cons
- ✗ No NVLink support (multi-GPU limited to PCIe P2P)
- ✗ Very expensive ($45,000)
- ✗ Very expensive ($45k+); large physical footprint; high power requirements
Compare with Other Workstations
Apple Mac Studio M2 Ultra (192GB)
Apple Mac Studio M2 Ultra (64GB)
ArsenalPC MES2X Dual RTX 5090 AI Workstation
BIZON G3000 G2 (1x RTX 5090)
BIZON G3000 G2 (4x RTX 5090)
BOXX APEXX 8R (1x RTX PRO 6000 Blackwell)
Custom Dual RTX 4090 Training Workstation
Custom RTX 5090 AI Workstation (Value Build)
Sources & Verification
Primary source: HP Z8 Fury G5 Datasheet
https://www.hp.com/us-en/workstations/z8-fury.html
Verification date: 2026-07-23
Data quality: Vendor specification — from manufacturer datasheet. No hands-on testing was performed.