NVIDIA RTX Pro 6000 Blackwell Workstation 96GB GDDR7 Graphics Card
MPN 900-2G153-0000-000 · Brand NVIDIA
The NVIDIA RTX Pro 6000 Blackwell Workstation GPU delivers 96 GB of GDDR7 memory at 1.8 TB/s bandwidth, 24,064 CUDA cores, 5th-Gen Tensor Cores with FP4, and PCIe Gen 5 — purpose-built for AI development, scientific simulation, and professional rendering workloads
Inclusive of GST · GST invoice available
Overview
The NVIDIA RTX Pro 6000 Blackwell Workstation Edition is a professional graphics card built on the NVIDIA Blackwell architecture, designed for AI developers, engineers, VFX artists, and researchers who require extreme compute throughput at the desktop workstation level. At its core are 24,064 CUDA parallel processing cores paired with 5th-generation Tensor Cores that deliver FP4 precision support, enabling faster AI model processing with reduced memory overhead — making local fine-tuning of large language models a practical workstation task. The 4th-generation Ray Tracing Cores double the ray-triangle intersection throughput compared to the previous generation, supporting RTX Mega Geometry for up to 100x more ray-traced triangles in complex scenes. With 96 GB of GDDR7 memory across a 512-bit interface, the card delivers up to 1.8 TB/s of memory bandwidth, giving it the headroom to handle massive 3D datasets, billion-parameter AI models, and simultaneous multi-application workflows without bottleneck. PCIe Gen 5 x16 connectivity doubles the host-to-GPU bandwidth versus PCIe Gen 4, keeping data pipelines open for demanding AI and data-science tasks. Video workflows benefit from 9th-generation NVENC and 6th-generation NVDEC engines with full 4:2:2 encoding and decoding support, accelerating high-resolution production pipelines. Four DisplayPort 2.1b connectors support resolutions up to 8K at 240 Hz or 16K at 60 Hz across up to four simultaneous displays. Multi-Instance GPU (MIG) support allows a single card to be partitioned into isolated instances with dedicated resources, enabling concurrent multi-tenant or multi-workload deployments. The double-flow-through thermal design channels airflow along dual axial pathways to sustain stable operation under the card's 600 W power envelope without thermal throttling. Verified Specifications: • Architecture: NVIDIA Blackwell • CUDA Cores: 24,064 • Memory: 96 GB GDDR7 • Memory Interface: 512-bit • Memory Bandwidth: 1.8 TB/s • Tensor Cores: 5th Generation (FP4 support) • RT Cores: 4th Generation • Video Engines: 9th-Gen NVENC, 6th-Gen NVDEC (4:2:2) • Display Outputs: 4 × DisplayPort 2.1b • Max Resolution: 7680 × 4320 • PCI Express: Gen 5.0 x16 • Power Connector: 1 × 16-pin (PCIe CEM5) • Recommended PSU: 600 W • OpenGL: 4.6 | DirectX: 12 • MIG Support: Yes • Warranty: 3 Years • SKU: 900-2G153-0000-000
Datasheet
| Memory Type | GDDR7 |
|---|---|
| Warranty | 3 Years |
| Memory Size | 96 GB |
| GPU | NVIDIA RTX Pro |
| Recommended PSU | 600 W |
| Resolution | 7680 x 4320 |
| Model | 900-2G153-0000-000 |
| Ports | Mini DisplayPort |
| Note** | Features, Price and Specifications are subject to change without notice. Images use for Representation Purpose Only |
| Opengl | 4.6 |
| Directx | 12 |
| Pci Express | 5.0 |
| Memory Interface | 512-BIT |
| Power Connectors | 1 x 16-pin |
| Max Display Support | 4 |
| GPU Core (Cuda Core) | 24064 |
| Cooler | Passive |
Complete your build




