NVIDIA RTX PRO 4000 Blackwell 24GB Workstation Graphics Card
MPN RTX-PRO-4000-BWL · Brand Nvidia · 2.7” x 6.6” L, dual slot, half height
The NVIDIA RTX PRO 4000 Blackwell delivers 24 GB GDDR7 ECC memory, 8,960 CUDA cores, and fifth-generation Tensor Cores in a compact half-height, dual-slot form factor — purpose-built for AI, CAD, and professional rendering workloads
Inclusive of GST · GST invoice available
Overview
The NVIDIA RTX PRO 4000 Blackwell is a professional workstation graphics card built on NVIDIA's Blackwell GPU architecture. It is designed for demanding workflows including AI model development, CAD engineering, 3D rendering, and visual effects production. At the heart of the card are 8,960 NVIDIA CUDA cores paired with 24 GB of GDDR7 memory featuring ECC (Error-Correcting Code) support, ensuring data integrity under sustained compute loads. The 192-bit memory interface delivers 432 GB/s of memory bandwidth, enabling fast data throughput for large scene files, simulation datasets, and AI training tasks. The Blackwell architecture introduces fifth-generation Tensor Cores for accelerated AI inference and training, and fourth-generation RT Cores for real-time ray tracing in supported professional applications. Both features work alongside support for CUDA 12.8 and OpenCL 3.0 compute APIs, as well as DirectX 12, Shader Model 6.7, OpenGL 4.6, and Vulkan 1.4 graphics APIs. Connectivity is handled via a PCIe 5.0 x8 system interface, while display output is provided through four Mini DisplayPort 2.1b connectors. The card supports up to four 4K displays at 165 Hz or up to two 8K displays at 100 Hz simultaneously. Video encoding and decoding are managed by a ninth-generation NVENC engine and a sixth-generation NVDEC engine respectively. The RTX PRO 4000 Blackwell operates within a 70 W maximum power envelope and uses an active thermal solution, making it suitable for compact workstation builds. Its half-height, dual-slot form factor measures 2.7 inches by 6.6 inches, enabling deployment in small form factor and rack-mount systems where space is constrained. Verified Specifications: - GPU Architecture: NVIDIA Blackwell - CUDA Cores: 8,960 - Tensor Cores: Fifth generation - RT Cores: Fourth generation - GPU Memory: 24 GB GDDR7 with ECC - Memory Interface: 192-bit - Memory Bandwidth: 432 GB/s - System Interface: PCIe 5.0 x8 - Display Connectors: 4x Mini DisplayPort 2.1b - Max Displays: 4x 3840×2160 @ 165 Hz or 2x 7680×4320 @ 100 Hz - Video Engines: 1x NVENC (9th gen), 1x NVDEC (6th gen) - Max Power Consumption: 70 W - Form Factor: 2.7″ × 6.6″ L, dual slot, half height - Compute APIs: CUDA 12.8, OpenCL 3.0 - Graphics APIs: DirectX 12, Shader Model 6.7, OpenGL 4.6, Vulkan 1.4 - Warranty: 3-year manufacturer warranty
Datasheet
| Power Consumption | Max power consumption: 70 W |
|---|---|
| GPU Memory | 24 GB GDDR7 with ECC |
| Compute Apis | CUDA 12.8, OpenCL 3.0 |
| Tensor Cores | Fifth generation |
| Graphics Apis | DirectX 12, Shader Model 6.7, OpenGL 4.62 , Vulkan 1.42 |
| Video Engines | 1x NVENC (ninth generation) 1x NVDEC (sixth generation) |
| GPU Architecture | NVIDIA Blackwell |
| Memory Bandwidth | 432 GB/s |
| Memory Interface | 192 bi |
| System Interface | PCIe 5.0 x 8 |
| Thermal Solution | Active |
| Ray Tracing Cores | Fourth generation |
| Display Connectors | 4x Mini DisplayPort 2.1b |
| Nvidia® Cuda® Cores | 8960 |
| Max Simultaneous Displays | > 4x 3840×2160 @ 165 Hz > 2x 7680×4320 @ 100 Hz |
| Form Factor | 2.7” x 6.6” L, dual slot, half height |
Complete your build
