Powerful, efficient, and particularly compact – that is the Nvidia RTX PRO 4000 Blackwell graphics card. Thanks to the 24 GB of GDDR7 graphics memory and the powerful RT and Tensor cores, this graphics card is ideal for anyone seeking workstation performance in a small form factor.
- Blackwell architecture with optimal performance for rendering, simulation, and streaming
- 24 GB GDDR7 ECC memory for smooth work on the largest projects
- Fast and photorealistic ray tracing thanks to fourth-generation RT cores
- Fifth-generation Tensor cores for uncompromising AI workflows
- Compact single-slot form factor for small workstations or multi-GPU applications
Nvidia RTX PRO 4000 Blackwell: Powerful workstation GPU in a compact format
From everyday AI workflows with LLMs, generative AI, or model training to demanding rendering projects or even the occasional gaming session – the Nvidia RTX PRO 4000 offers more than enough raw power with its 8,960 CUDA cores, which, together with RT and Tensor cores, ensures your workday runs smoothly.
AI accelerators and ray tracing
Whether it’s a compute-intensive AI workflow or photorealistic ray-traced renderings – integrated fourth-generation RT cores work in harmony with fifth-generation Tensor cores to effectively accelerate your workflow with modern tools.
Exceptionally compact form factor
The RTX PRO 4000 offers another unique feature: the graphics card’s form factor. Thanks to its single-slot design, the card is especially suitable for smaller workstations. But even in larger cases or rack mounts, the graphics card benefits from the single-slot design, allowing multiple graphics cards to fit uncompromisingly on the same motherboard.
|
Colour
|
|
|
Primary Colour
|
Black
|
|
Dimensions
|
|
|
Length / Depth
|
241 mm
|
|
Width
|
111 mm
|
|
Height
|
20 mm
|
|
Graphics Card
|
|
|
GPU Series
|
RTX PRO
|
|
GPU Model
|
NVIDIA RTX Pro 4000 Blackwell Workstation Edition
|
|
GPU Series Name
|
Blackwell PRO W
|
|
Graphics Chip
|
GB203-400
|
|
Manufacturing Process
|
5 nm
|
|
Shader Units
|
8960
|
|
Tensor Cores
|
280
|
|
Raytracing Cores / Ray Acceleratorss
|
70
|
|
GPU Memory Type
|
GDDR7
|
|
Slot Type Standard
|
PCIe 5.0
|
|
GPU Memory Size (GB)
|
24
|
|
GPU Memory Interface (bit)
|
192
|
|
GPU Memory Bandwidth (GB/s)
|
672
|
|
GPU Memory Clock (MHz)
|
1750
|
|
GPU Memory Clock Effective (Gbps)
|
28
|
|
Clock Speeds
|
|
|
Max. GPU Clock (Base)
|
1230 MHz
|
|
Max. GPU Clock (Boost)
|
2055 MHz
|
|
Cooling
|
|
|
Number of GPU Fans
|
1
|
|
Water Cooling
|
|
|
Watercooling
|
No
|
|
Internal Ports
|
|
|
Required Power Connectors
|
16-Pin PCIe 5.0 - 12v-2x6 (1x)
|
|
External Ports
|
|
|
Display Connectors
|
Display Port (4x)
|
|
Display Connector Standards
|
DisplayPort 2.1
|
|
Back Connect
|
|
|
Back Connect
|
No
|
|
Expansion Slots
|
|
|
Expansion Slots Required
|
1 Slot
|
|
Standards / Specifications
|
|
|
Adaptive Sync Technology (G-SYNC / FreeSync)
|
NVIDIA G-SYNC
|
|
Compatibility
|
|
|
Compatibility Note
|
Please refer to the manufacturer's compatibility information. You can find the latest overview on the manufacturer's website.
|
|
AI
|
|
|
AI Acceleration
|
Yes
|
|
AI Accelerator
|
Tensor Cores
|
|
Lighting
|
|
|
Lighting / RGB
|
No
|
Product reviews displayed are aggregated from all Pro Gamers Group websites.
| Manufacturer Information | |
| Responsible Person |