Ultra 9 AI Workstation with RTX PRO 6000 Hand-built · burn-in tested
NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96 GB GPU memory · 4,000 AI TOPS
AI WORKSTATIONIn stock · ships in 7–10 business days

Ultra 9 AI Workstation with RTX PRO 6000

96 GB of GPU memory on a single card. Run 70B-class models and gpt-oss-120b entirely on the GPU, privately, on your desk.

Best forLocal LLMs up to 120B70B models at 8-bitFine-tuning larger models
GraphicsRTX PRO 6000 · 96 GB
ProcessorCore Ultra 9 285K
Memory64 GB DDR5-5600
Storage2 TB PCIe 5.0 NVMe SSD (9100)
$28,799Full specs
Buying for a business? Get a quote, PO or net terms
  • 48-hour burn-in test
  • 3-year limited warranty
  • Free shipping
  • Lifetime tech support
RTX PRO 6000$28,799
RTX PRO 6000 workstation$28,799

Overview

Local AI, without the cloud bill

The Ultra 9 AI Workstation with RTX PRO 6000 pairs NVIDIA's 96 GB Blackwell workstation GPU with Intel's 24-core Core Ultra 9 285K. That's three times the GPU memory of our RTX 5090 build: enough for 70B-class language models at 8-bit and OpenAI's gpt-oss-120b, entirely on the GPU.

Inside a Fractal Design Define 7 you get 64 GB of DDR5-5600, a 2 TB PCIe 5.0 NVMe SSD, a be quiet! Silent Loop 3 liquid cooler and a 1000 W 80 PLUS Gold power supply.

96 GBGPU memorySets which AI models fit on the RTX PRO 6000.
4,000AI TOPSNVIDIA's rating for AI throughput.
24 coresCore Ultra 9 285KFor data prep, builds and everything else.
64 GBSystem memory64 GB DDR5-5600

What's inside

What's inside the RTX PRO 6000 workstation

Every part is listed by name, so you know exactly what you're getting and what you can upgrade later.

1 2 3 4 5 6 7
  1. 1
    GraphicsNVIDIA RTX PRO 6000 Blackwell Workstation Edition96 GB GPU memory · 4,000 AI TOPS
  2. 2
    ProcessorIntel Core Ultra 9 285K24 cores (8 Performance + 16 Efficient)
  3. 3
    CPU coolerbe quiet! Silent Loop 3
  4. 4
    Memory64 GB DDR5-5600
  5. 5
    Storage2 TB PCIe 5.0 NVMe SSD (9100)
  6. 6
    Power supplyGigabyte GP-UD1000GM PG5 V2 · 1000 W 80 PLUS Gold · ATX 3.1
  7. 7
    CaseFractal Design Define 7 mid-tower

Use cases

What you can do with it

Everything runs on the machine under your desk: no per-hour bill, no usage caps, and nothing you type or upload leaves the building.

Running locally

A private AI assistant

Run open models like gpt-oss, Qwen and Llama for your team. Prompts and data stay on your network.

contract.pdf Page 4

Chat with your documents

Point a local model at contracts, manuals or code and ask questions, without uploading files anywhere.

Generate

Generate images

Create product shots, concept art and marketing visuals with Stable Diffusion and FLUX.

Training loss LoRA adapter

Fine-tune and build

Train LoRA adapters on your own data and prototype AI apps with CUDA, PyTorch and Docker.

Model fit

What AI models can the RTX PRO 6000 workstation run?

GPU memory (VRAM) decides which models fit. This build has 96 GB, so every bar that ends left of the line runs fully on the GPU.

Fits in GPU memory Needs more memory This build: 96 GB
8B models · 8-bitLlama 3.1 8B, Qwen3 8B
~10 GB
Runs on this buildRuns
12–14B models · 4-bitGemma 3 12B, Qwen3 14B
~11 GB
Runs on this buildRuns
SDXL image generationStable Diffusion XL
~12 GB
Runs on this buildRuns
gpt-oss-20bOpenAI open-weight model
~14 GB
Runs on this buildRuns
27B models · 4-bitGemma 3 27B
~19 GB
Runs on this buildRuns
32B models · 4-bitQwen3 32B
~22 GB
Runs on this buildRuns
FLUX.1 [dev] · full precisionFLUX image model
~28 GB
Runs on this buildRuns
70B models · 4-bitLlama 3.3 70B
~44 GB
Runs on this buildRuns
gpt-oss-120bOpenAI open-weight model
~66 GB
Runs on this buildRuns
70B models · 8-bitLlama 3.3 70B
~80 GB
Runs on this buildRuns
70B models · full precisionLlama 3.3 70B
~150 GB
Ask about multi-GPU buildsAsk us

Approximate memory for the model plus a working context. Larger models can run partly on the CPU and system memory, much more slowly.

Compare

Compare the AI workstation lineup

Same case, cooler and build process. The GPU sets which models you can run; the CPU and memory set how much else it handles at once.

RTX 5070 Ti$3,999View build
RTX 5080$6,499View build
RTX 5090$11,499View build
YOU'RE VIEWINGRTX PRO 6000$28,799
GPU memory16 GB16 GB32 GB96 GB
AI TOPS1,4061,8013,3524,000
ProcessorCore Ultra 7 265KCore Ultra 9 285KCore Ultra 9 285KCore Ultra 9 285K
CPU cores20242424
System memory32 GB DDR5-560064 GB DDR5-560064 GB DDR5-560064 GB DDR5-5600
Storage2 TB PCIe 5.0 NVMe SSD (9100)2 TB PCIe 5.0 NVMe SSD (9100)2 TB PCIe 5.0 NVMe SSD (9100)2 TB PCIe 5.0 NVMe SSD (9100)
Power supplyGigabyte GP-UD850GM PG5 V2 · 850 W 80 PLUS Gold · ATX 3.1Gigabyte GP-UD1000GM PG5 V2 · 1000 W 80 PLUS Gold · ATX 3.1Gigabyte GP-UD1000GM PG5 V2 · 1000 W 80 PLUS Gold · ATX 3.1Gigabyte GP-UD1000GM PG5 V2 · 1000 W 80 PLUS Gold · ATX 3.1

Support

Built and backed by people you can call

Burn-in report Ultra 9 AI Workstation with RTX PRO 6000 GPU stressCPU stressMemory testStorage check PASS PASS PASS PASS GPU temperature Start End of test

48-hour burn-in

Every system is assembled and stress-tested in the USA for 48 hours before it ships.

Terminal $ nvidia-smi --query-gpu=name --format=csv,noheader NVIDIA RTX PRO 6000 Blackwell Workstation Edition $ python3 -c "import torch; print(torch.cuda.is_available())" True $ ollama run gpt-oss:20b >>> Send a message (/? for help)

Ready on first boot

Windows 11 Pro with current NVIDIA drivers, set up and tested before it ships.

3-year limited warranty

Hardware coverage handled by the team that built your system.

Lifetime technical support

Questions about drivers, models or upgrades go straight to our build team.

Specifications

Full specifications

Ultra 9 AI Workstation with RTX PRO 6000

Compute

Processor
Intel Core Ultra 9 285K24 cores (8 Performance + 16 Efficient)
Graphics
NVIDIA RTX PRO 6000 Blackwell Workstation Edition24,064 CUDA cores · 96 GB GDDR7 with ECC · 1,792 GB/s memory bandwidth · 4,000 AI TOPS
GPU memory
96 GB

Memory and storage

System memory
64 GB DDR5-5600
Storage
2 TB PCIe 5.0 NVMe SSD (9100)

Platform

Motherboard
Z890 Eagle
Power supply
Gigabyte GP-UD1000GM PG5 V2 · 1000 W 80 PLUS Gold · ATX 3.1
CPU cooler
be quiet! Silent Loop 3
Case
Fractal Design Define 7 mid-tower

Software and connectivity

Operating system
Windows 11 Pro

Size, power and warranty

Warranty
3-year limited hardware warranty

FAQ

Questions before you buy

Can't find your answer? Contact us.

What can I run on it?

Check the model list above: anything marked “Runs on this build” fits fully in GPU memory. Working with something that isn't listed? Ask us before you buy.

Which operating system does it come with?

Windows 11 Pro, set up with current NVIDIA drivers and tested before it ships.

Does it need special power or cooling?

No. Every build uses an 80 PLUS Gold power supply (850 W or 1000 W, depending on the GPU) and runs from a standard household outlet. The CPU is cooled by a be quiet! Silent Loop 3 liquid cooler inside a sound-dampened Fractal Design Define 7 case.

Can I upgrade it later?

Yes, but contact us first. Hardware installed after shipment isn't covered by the warranty, and we can confirm which parts are compatible.

How long until it ships?

Each system is built to order, burn-in tested for 48 hours and ships in 7–10 business days. Shipping is free.

Do you sell to businesses?

Yes. We accept purchase orders and GSA, offer net terms, and quote multi-unit orders. Request a quote.

Ready to order?

$28,799

Ships in 7–10 business days

Need a different configuration?

Multi-GPU builds, more memory, rack-mount servers, or several units for a team. Tell us the workload and we'll send a quote.

Request a quote