Two Flagships. One Box.
The top rung of all eight configurations on this listing: the RTX PRO 6000 Blackwell at its full 600 W GPU power budget, and the Ryzen 9950X3D at up to 5.7 GHz — because real AI work is a duet of GPU serving and CPU iteration, and this box refuses to pick one. OpenClaw ships pre-installed on Ubuntu.
The Two Chips Doing the Heavy Lifting
Hover each card — then read what the numbers actually buy your AI workloads:
RTX PRO 6000 Blackwell · 600 W GPU
- 96 GB of GDDR7 on one card — the largest open-weight models this family can run, loaded whole, with context to spare. No sharding, no cross-card complexity.
- ECC video memory — weights on a weeks-long service can't be silently corrupted by a bit-flip. Consumer cards don't do this.
- Full 600 W GPU budget — sustained tokens-per-second (the power-tuned MaxQ sibling is on this listing if you'd rather trade throughput for quiet).
- Universal MIG — partition the flagship into isolated instances and share it across researchers or services, safely.
AMD Ryzen 9950X3D · Liquid-Cooled
- 128 MB of 3D V-Cache — tokenizers, dataloaders, and hot working sets stay on-die; every code-run-inspect lap gets shorter.
- 5.7 GHz max turbo — single-thread latency is where agent decisions, eval scripts, and notebook cells live. You feel this one daily.
- 16 cores — preprocessing, indexing, and OpenClaw agent orchestration run beside the GPU workload without queueing.
- ECC UDIMM platform + liquid loop — research-grade data integrity with clocks that hold overnight, quietly.
Manufacturer-Published, Footnoted, No Hand-Waving
The Back Tells the Truth. Hover the Dots.
1
Three networks, three jobs1× 10 Gb/s for datasets, 2.5 Gb/s for the office LAN, plus a dedicated management port for IT.
2
Full-mesh exhaustThe 600 W GPU breathes through mesh on every panel that matters.
3
Flagship GPU outputsThe RTX PRO 6000 Blackwell's display outputs — color-accurate multi-monitor for the creation half of your day.
4
1300 W 80 PLUS PlatinumSized so the full-power flagship never negotiates for watts.
- 1
Three networks, three jobs
1× 10 Gb/s moves datasets at wire speed, 2.5 Gb/s serves the office, and the management port lets IT run it like server gear.
- 2
Full-mesh exhaust path
A 600 W GPU is an airflow problem; this chassis treats it like one.
- 3
Flagship display outputs
Drive color-accurate multi-monitor setups for rendering and review.
- 4
1300 W 80 PLUS Platinum
Premium-efficiency headroom for both flagships at full song.
Days This Box Was Built For
The Biggest Models, No Waiting List
96 GB of ECC GDDR7 loads the largest open-weight models this family can run — long context, full precision options, zero cloud queue. Your data never leaves the building.
Rendering by Day, Generation by Night
The same silicon that serves models also ray-traces scenes and drives generative pipelines — while the X3D CPU keeps viewports, timelines, and notebooks feeling instant.
Production Inference That Doesn't Blink
Full 600 W GPU sustained throughput, ECC end to end, and OpenClaw agents pre-installed for the operational glue — a private endpoint that treats 3 a.m. like 3 p.m.
Scene photos are AI-generated illustrations of typical deployments, not photographs of this exact product.
Everything Behind the Two Thrones
| ZRR9950X3D-RP6000 · As Configured | |
|---|---|
| GPU | RTX PRO 6000 Blackwell — 600 W GPU power budget, 96 GB GDDR7 with ECC |
| CPU | AMD Ryzen 9950X3D — 16-core, liquid-cooled, 4.3 GHz base / 5.7 GHz max turbo, 128 MB L3 (3D V-Cache), Socket AM5 |
| Memory | 128 GB DDR5 ECC UDIMM — expandable to 192 GB across 4 DIMM slots |
| Storage | 2× 2 TB M.2 NVMe (4 TB total) |
| Network | 1× 10 Gb/s + 2.5 Gb/s LAN + dedicated management port |
| Expansion | PCIe 5.0 x16 + PCIe 4.0 x16 · ASUS Pro WS B850M-ACE SE |
| Power | 1300 W 80 PLUS Platinum |
| OS | Ubuntu pre-installed, OpenClaw included (Windows available by quote) |
Eight Configs. This Is the Peak.
RTX PRO 4500 tier — LLM apps, RAG, and agents.
RTX PRO 5000 tier — mid-size model serving.
Full 600 W GPU with the 9950X3D. A MaxQ variant and EPYC 4565P CPU alternatives are also on this listing — compare options on this page.
Quick Answers Before You Buy
Why pair the 96 GB GPU with the X3D CPU?
Real AI days alternate: the GPU serves and trains while the CPU runs notebooks, evals, tokenizers, and agents. This config removes the VRAM ceiling and the iteration lag at the same time.
This vs. the MaxQ variant?
Both carry 96 GB. This runs the full 600 W GPU budget for maximum sustained throughput; MaxQ is power-tuned. The 1300 W 80 PLUS Platinum supply here is sized for full power.
Can several people share it?
Yes — Universal MIG partitions the GPU into isolated instances, served over 1× 10 Gb/s + 2.5 Gb/s LAN with a dedicated management port for IT.
What does "OpenClaw pre-installed" mean?
The OpenClaw enterprise AI agent framework comes ready on Ubuntu — agentic workflows on day one, no setup. Windows is available by quote.
Government / public-sector eligible?
TAA compliant and assembled in the USA, with compliance documents available for purchase review.
How is it supported?
1-year parts and labor limited warranty backed by ABS manufacturer support. Build to Order — custom built and stress-tested after purchase.
