NVIDIA RTX PRO 6000 Blackwell Server Edition 96GB GDDR7 (Bulk)
NVIDIA RTX PRO 6000 Blackwell Server Edition GPU with 96GB GDDR7 ECC memory, 120 TFLOPS FP32, PCIe 5.0 x16. Passive cooled, dual-slot for AI, rendering, scientific computing. Up to 600W, 4x DisplayPort 2.1.
Please Note
The website is currently being actively developed, you are welcome to place the order and in 99% of the cases your order will be processed as usual, but in some cases there may be a delay or the price may change. We will notify you if there are any issues.
Discontinued
This item was discontinued and currently is not available for order on our website. The item may become available soon. Please contact us if you need more information regarding the upcoming arrivals.
| GPU (Graphics Cards) - Series | NVIDIA RTX Professional |
|---|---|
| GPU (Graphics Cards) - PCIe Interface | PCIe 5.0 x16 |
| GPU (Graphics Cards) - Cooler Type | Passive |
| GPU (Graphics Cards) - VRAM Capacity | 96 GB |
| GPU (Graphics Cards) - VRAM Type | GDDR7 |
| GPU (Graphics Cards) - Outputs | DisplayPort 2.1 |
| GPU (Graphics Cards) - Card Length | 360 mm+ |
The NVIDIA RTX PRO 6000 Blackwell Server Edition is a high-performance data center GPU powered by the cutting-edge Blackwell architecture. Equipped with 96GB of GDDR7 memory with ECC, it delivers exceptional capabilities for multimodal AI inference, physical AI, scientific computing, rendering, 3D graphics, and video processing. Featuring fifth-generation Tensor Cores with FP4 precision, fourth-generation RT Cores, and up to 120 TFLOPS FP32 performance, this GPU accelerates complex workflows in industries like engineering, healthcare, and media.
Designed for server environments, it uses a passive cooling solution in a dual-slot, full-height/full-length form factor (4.4" H x 10.5" L). Supports PCIe Gen5 x16 interface, multi-instance GPU (MIG) up to 4x24GB, confidential computing, and 4x NVENC/NVDEC engines. Power configurable up to 600W, ideal for dense computing nodes.
- 96GB GDDR7 ECC memory, 512-bit, 1597 GB/s bandwidth
- 24,064 CUDA cores, 752 Tensor/188 RT cores
- 4x DisplayPort 2.1 outputs
- PCIe 5.0 x16, passive cooled, server-optimized
- Bulk packaging for enterprise deployment
| GPU Architecture | NVIDIA Blackwell |
| CUDA Cores | 24,064 |
| Tensor Cores | 752 (5th Gen, FP4) |
| RT Cores | 188 (4th Gen) |
| FP32 Performance | 120 TFLOPS |
| RT Core Performance | 355 TFLOPS |
| FP4 AI Performance | 4 PFLOPS |
| GPU Memory | 96 GB GDDR7 with ECC |
| Memory Interface | 512-bit |
| Memory Bandwidth | 1597 GB/s |
| Graphics Bus | PCIe 5.0 x16 |
| Display Outputs | 4x DisplayPort 2.1 |
| Power Consumption | Up to 600W (configurable) |
| Form Factor | Dual-slot, 4.4 |
| Cooler Type | Passive |
| Multi-Instance GPU | Up to 4 MIGs @ 24GB |
| NVENC/NVDEC | 4x / 4x (9th/6th Gen) |