.png?width=274&height=200&scale=canvas)
.png?width=500?width=274&height=200&scale=canvas)
.png?width=274&height=200&scale=canvas)
.png?width=274&height=200&scale=canvas){kind=icon}
.jpg)
.jpg)
.jpg)
.jpg)
NVIDIA RTX PRO 5000 Blackwell 72GB GDDR7 Graphics Card
Available for delivery from Thursday 9th July
Click & Collect from Tomorrow
Low stock - don't miss out
Order within
You have 0 of these in your basket
The RTX PRO 5000 Blackwell is NVIDIA's high-end professional workstation GPU aimed at AI development, local LLMs, engineering, CAD, scientific computing, rendering, and content creation. It sits between the RTX PRO 4500 and the RTX PRO 6000 Blackwell in NVIDIA's workstation lineup.
Key Selling Points
-
Massive Professional AI Performance
Built on the NVIDIA Blackwell architecture, the RTX PRO 5000 features 5th Generation Tensor Cores, 4th Generation RT Cores, and up to 2,064 AI TOPS, making it ideal for local AI assistants, LLM inference, generative AI, and AI-enhanced creative workflows. NVIDIA states Tensor Core performance is up to 3× higher than the previous generation in supported workloads.
-
48GB or 72GB ECC GDDR7 Memory
Available in both 48GB ECC GDDR7 and 72GB ECC GDDR7 variants, the card provides up to 1,344 GB/s memory bandwidth, allowing professionals to work with larger AI models, more complex simulations, massive datasets, and advanced 3D scenes directly on a workstation.
-
Workstation-Optimised Design
Unlike gaming-focused RTX cards, the RTX PRO 5000 includes professional drivers, ECC memory, PCIe 5.0 support, enterprise certifications, Multi-Instance GPU (MIG) support, and a dual-slot active cooling design suitable for multi-GPU workstations.
Key Specifications
| Specification | RTX PRO 5000 Blackwell |
|---|---|
| Architecture | NVIDIA Blackwell |
| AI Performance | 2,064 AI TOPS |
| Tensor Cores | 5th Generation |
| RT Cores | 4th Generation |
| Memory | 48GB or 72GB ECC GDDR7 |
| Memory Bandwidth | 1,344 GB/s |
| PCIe Interface | PCIe 5.0 x16 |
| Display Outputs | 4 × DisplayPort 2.1 |
| NVENC | 3 × 9th Gen |
| NVDEC | 3 × 6th Gen |
| Power Consumption | 300W |
| Form Factor | Dual-Slot Active Cooling |
RTX PRO 5000 vs RTX 5090
| Feature | RTX PRO 5000 | RTX 5090 |
|---|---|---|
| VRAM | 48GB / 72GB ECC | 32GB GDDR7 |
| ECC Memory | Yes | No |
| Power Draw | 300W | 575W |
| Professional Drivers | Yes | No |
| AI / CAD Certification | Yes | No |
| Gaming Performance | Lower | Higher |
| Multi-GPU Workstations | Excellent | Limited |
For AI development, CAD, simulation, rendering, and workstation use, the RTX PRO 5000 is generally the better choice. For pure gaming, the RTX 5090 remains substantially faster and far more cost-effective. Community discussions around the 72GB version have focused heavily on its appeal for local AI workloads thanks to the large VRAM capacity.
GPU
- Chipset Manufacturer: NVIDIA
- Chipset Model Name: RTX PRO 5000 Blackwell
- Number of Cores: CUDA: 14,080
- Tensor: 5th Gen
- Ray Tracing: 4th Gen
- Cooling System: Active
Memory
- Memory Size: 72 GB with ECC
- Memory Type: GDDR7
- Memory Width: 512-bit
- Memory Bandwidth: 1.344 GB/s
Interfaces/Ports
- Video Component Ports: PCIe 5.0 x16
- DisplayPort: 4x DP v2.1b
Performance
- Floating Point Operations (Single): AI TOPS: 2064 AI TOPS
- 65 TFLOPS
- Floating Point Operations (Double): 196 TFLOPS
Resolution
- Maximum Resolution: 7680 x 4320 @ 60 Hz
System
- Driver Support: Graphics APIs: DirectX 12, Shader Model 6.6, OpenGL 4.6, Vulkan 1.3
- Compute APIs: CUDA 12.8, OpenCL 3.0, DirectCompute
Electrical
- Power Connector: 1x PCIe CEM5 16-pin
- Power Consumption: TDP 300w
Physical
- Width: 111.76 mm
- Depth: 266.7 mm
- Height: Dual Slot
Identifiers
- Stock Code: VGA7610
- Mfg Code: 900-5G153-2270-000
.jpg?width=2000)