NVIDIA RTX PRO 5000 Blackwell GPU

NVIDIA RTX PRO 5000 Blackwell GPU
NVIDIA RTX PRO 5000 Blackwell GPUNVIDIA RTX PRO 5000 Blackwell GPUNVIDIA RTX PRO 5000 Blackwell GPUNVIDIA RTX PRO 5000 Blackwell GPUNVIDIA RTX PRO 5000 Blackwell GPUNVIDIA RTX PRO 5000 Blackwell GPUNVIDIA RTX PRO 5000 Blackwell GPU

NVIDIA RTX PRO 5000 Blackwell GPU

$3,118.80$7,797.00

The NVIDIA RTX PRO 5000 Blackwell is a desktop GPU built on the NVIDIA Blackwell architecture with 48 GB of GDDR7 memory and ECC, designed for generative AI, 3D rendering, and real-time simulation. Its 48 GB capacity is enough to run Llama 3.3 70B fully in VRAM, serve 4–6 users from one card via MIG, and fine-tune 34B models without offloading to system RAM.

Native FP4 support on 5th-generation Tensor Cores doubles effective memory versus FP8 workflows, while ECC memory protects long training runs from silent bit-flips. The card connects over PCIe 5.0 x16 and provides four DisplayPort 2.1b outputs, with 2x NVENC (9th Gen) and 2x NVDEC (6th Gen) video engines for multi-stream work.

Key Features

Brand NVIDIA
Model RTX PRO 5000 Blackwell
Product Type Graphics Card
GPU Architecture NVIDIA Blackwell
CUDA Cores 14,080
Tensor Cores 5th Generation
Ray Tracing Cores 4th Generation
GPU Memory 48 GB GDDR7 with ECC
Memory Interface 384-bit
Memory Bandwidth 1344 GB/s
System Interface PCIe 5.0 x16
Display Connectors 4x DisplayPort 2.1b
Max Simultaneous Displays 4x 3840 x 2160 @ 165 Hz, 2x 7680 x 4320 @ 100 Hz
Video Engines 2x NVENC (9th Gen), 2x NVDEC (6th Gen)
MIG Instance Types Up to 2x 24 GB, Up to 1x 48 GB
Power Consumption 300 W
Power Connector 1x PCIe CEM5 16-pin
Thermal Solution Active
Form Factor 4.4" x 10.5" L, dual slot, full height
Graphics APIs DirectX 12, Shader Model 6.6, OpenGL 4.63, Vulkan 1.3
Compute APIs CUDA 12.8, OpenCL 3.0, DirectCompute
Product Weight 4 lb

Highlights

  • 48 GB of GDDR7 memory. Enough to run Llama 3.3 70B fully in VRAM, serve 4–6 users from one card via MIG, and fine-tune 34B models without offloading to system RAM.
  • Native FP4 on 5th-gen Tensor Cores doubles effective memory versus FP8 workflows.
  • ECC memory protects long training runs from silent bit-flips.

Compatibility

Requires a PCIe 5.0 x16 slot and a single PCIe CEM5 16-pin power connector. Supports DirectX 12, Shader Model 6.6, OpenGL 4.63, Vulkan 1.3, CUDA 12.8, OpenCL 3.0, and DirectCompute.

Warranty

Extended 3-Year Warranty

Frequently asked questions

What is the total board power and power connector of the RTX PRO 5000 Blackwell?

The total board power is 300 W, and it uses a single PCIe CEM5 16-pin power connector.

What display outputs and maximum resolutions does the card support?

It has four DisplayPort 2.1b connectors and supports up to 4x 3840 x 2160 @ 165 Hz or 2x 7680 x 4320 @ 100 Hz displays simultaneously.

What is the physical size and slot width of the RTX PRO 5000 Blackwell?

The card measures 4.4" x 10.5" L and occupies a dual-slot, full-height form factor.

What graphics and compute APIs are supported?

It supports DirectX 12, Shader Model 6.6, OpenGL 4.63, and Vulkan 1.3 for graphics, plus CUDA 12.8, OpenCL 3.0, and DirectCompute for compute.