Home / NVIDIA GPUs / RTX PRO 6000 96GB
Blackwell · 96GB GDDR7 · Three editions
NVIDIA RTX PRO 6000 Blackwell 96GB GPU, in Egypt & the UAE.
The flagship RTX PRO card: 96GB of GDDR7 and 24,064 CUDA cores for AI inference, fine-tuning, rendering and simulation. Available as the Workstation, Max-Q and Server Editions, factory-new with a 12-month replacement warranty.
96GB GDDR7 with ECC
512-bit memory with up to 1.8 TB/s of bandwidth.
24,064 CUDA cores
752 fifth-generation Tensor Cores and 188 fourth-generation RT Cores.
Three editions
Workstation (600W), Max-Q (300W) and the passive Server Edition.
12-month replacement
Factory-new, fit-checked by our engineers, delivered in Egypt and the UAE.
What it’s for
One card for AI and graphics.
With 96GB on one card, models that needed two GPUs now fit on one, and the same card still renders and drives displays.
Run 70B models on one GPU
A 70B-parameter model at FP8 fits in 96GB, and OpenAI’s gpt-oss-120b runs on a single card.
Fine-tune and prototype
Fine-tune open models on your own hardware, without sending data to the cloud.
AI video and images
Text-to-video and image generation, with 4 NVENC and 4 NVDEC engines per card.
Rendering and digital twins
188 RT Cores for ray-traced rendering, VFX and visualisation.
Simulation and engineering
Up to 125 TFLOPS of FP32 for engineering simulation and scientific work.
One card, several teams
Split each card into up to four isolated 24GB MIG instances.
Editions
Three editions of the same GPU.
Workstation Edition
600W with double flow-through fans. The fastest edition, for one or two cards in a desk-side workstation.
Max-Q Workstation Edition
300W with a blower that exhausts out of the back, so two to four cards can sit side by side.
Server Edition
Passive and configurable from 400W to 600W, for 4 to 8 cards in a rack server, like our RTX PRO 6000 8-GPU server.
Specifications
RTX PRO 6000 Blackwell specifications.
NVIDIA’s published specifications. All three editions share the same GPU and 96GB of memory; power, cooling and clock speeds differ.
GPU: RTX PRO 6000 Blackwell
| Architecture | NVIDIA Blackwell |
|---|---|
| CUDA cores | 24,064 |
| Tensor Cores | 752 (5th generation) |
| RT Cores | 188 (4th generation) |
| GPU memory | 96GB GDDR7 with ECC |
| Memory interface | 512-bit |
| Memory bandwidth | Up to 1.8 TB/s |
| FP32 | Up to 125 TFLOPS* |
| AI performance | Up to 4,000 AI TOPS* |
| Multi-Instance GPU | Up to 4 instances @ 24GB |
| Encoders / decoders | 4 NVENC (9th gen), 4 NVDEC (6th gen) |
| Display outputs | 4 × DisplayPort 2.1b |
| Interface | PCIe 5.0 x16 |
* Workstation Edition peak figures; AI TOPS are FP4 with sparsity. The Max-Q and Server Editions run at lower power and peak figures.
Editions compared
| Workstation Edition | 600W · double flow-through fans · dual-slot · 5.4" × 12" |
|---|---|
| Max-Q Workstation Edition | 300W · blower, rear exhaust · dual-slot · 4.4" × 10.5" |
| Server Edition | 400–600W configurable · passive, cooled by the server · dual-slot · 4.4" × 10.5" |
| Display outputs | On for the Workstation and Max-Q Editions; off by default on the Server Edition |
| Warranty | 12-month Mega Build Up replacement warranty |
Not sure which edition fits? Tell us your workstation or server model and we’ll check.
How we deliver
From first question to a working GPU.
01
Tell us your system
Share the server or workstation model and what you want to run on it.
02
We check the fit
Slots, power connectors, power supply, airflow and drivers, before you order.
03
Source factory-new
New, genuine NVIDIA GPUs at today’s price.
04
Deliver or install
Delivered across Egypt and the UAE, or installed and tested by our engineers.
05
12-month replacement
If a card fails within 12 months, we replace it.
Compare NVIDIA GPUs
Four GPUs, from 32GB to 141GB.
Efficient AI and design
RTX PRO 4500 Blackwell
32GB GDDR7
Up to 896 GB/s · 200W workstation or 165W server card
Large models on a desk
RTX PRO 5000 Blackwell
72GB GDDR7
1,344 GB/s · 300W blower · AI workstations
Flagship RTX PRO
RTX PRO 6000 Blackwell
96GB GDDR7
Up to 1.8 TB/s · workstations and servers
You are here
Which edition should I buy?
For one or two cards in a workstation, the Workstation Edition. For two to four cards side by side in a workstation, the Max-Q. For a rack server with 4 to 8 GPUs, the Server Edition, which is passive and relies on the server’s fans. We confirm what your system supports.
Which AI models fit on one card?
A 70B-parameter model at FP8 fits in 96GB with room for context, and OpenAI’s gpt-oss-120b runs on a single card. For training large models, the H200 NVL has 141GB of faster HBM3e memory and NVLink.
RTX PRO 6000 or H200 NVL?
The RTX PRO 6000 is the better all-rounder: AI inference, fine-tuning, rendering and video, with display outputs. The H200 NVL is a compute-only data-centre GPU, with about 2.7 times the memory bandwidth and NVLink, for training and the largest models.
Can I use it in a server I already have?
Often, yes. Tell us the server model and we check whether it supports the Server Edition, its 16-pin power connector and its airflow. If you need a new server, see our RTX PRO 6000 8-GPU server.
What warranty do the GPUs have?
Every GPU comes with a 12-month Mega Build Up replacement warranty. If a card fails, our engineers diagnose it and we replace it. See our warranty services.
Do you deliver and install in Egypt and the UAE?
Yes. We deliver across Egypt and the UAE, and our engineers can install the cards in your servers or workstations, update drivers and stress-test the system before handover.
How do I get a price?
GPU prices change often, so we quote today’s price. Message us on WhatsApp or answer three quick questions in the form on this page, and an engineer replies within 24 working hours with the price and availability of the RTX PRO 6000 edition you need.
Get a quote
Get today’s RTX PRO 6000 price.
Answer three quick questions and an engineer replies within 24 working hours with today’s price, availability and a fit check for your system.
- No obligation, no sales pressure
- Fit checked by an engineer before you order
- 12-month replacement warranty
Step 1 of 3
Talk to a GPU engineer.
Call, WhatsApp or email our team and we’ll reply within 24 working hours.
