Powered by the NVIDIA Blackwell architecture, it combines 96GB of GDDR7 ECC memory, fifth-generation Tensor Cores, fourth-generation RT Cores and PCIe Gen 5 connectivity to deliver exceptional performance for local AI development, large-model inference and professional visualization. NVIDIA positions the Workstation Edition specifically for maximum single-GPU workstation performance.

Key Specifications

• GPU Architecture: NVIDIA Blackwell
• CUDA Cores: 24,064
• GPU Memory: 96GB GDDR7 with ECC
• Memory Bandwidth: 1,792 GB/s
• AI Performance: Up to 4,000 AI TOPS
• Tensor Cores: 5th Generation
• RT Cores: 4th Generation
• Interface: PCIe 5.0 x16
• Display Outputs: 4 × DisplayPort 2.1b
• Power Consumption: 600W
• Cooling: Double Flow Through
• Form Factor: Dual-slot, 5.4″ × 12″

Built for Professional AI and GPU Workstations

With 96GB of ECC GPU memory, the RTX PRO 6000 Blackwell Workstation Edition is designed for workloads that exceed the memory capacity of conventional consumer GPUs.

Ideal applications include:

• Large language model inference
• Generative AI and AI agents
• Local AI development
• 3D rendering and visualization
• Engineering and CAD
• Scientific computing
• Simulation and digital twins
• NVIDIA Omniverse
• Video production
• Multimodal AI

Its fifth-generation Tensor Cores support FP4 AI workloads, while fourth-generation RT Cores accelerate professional ray tracing and neural rendering. NVIDIA

RTX PRO 6000 Workstation Edition vs Server Edition

The Workstation Edition is optimized for high-performance desktop and professional workstation systems and uses a double-flow-through cooling design.

The Server Edition is designed for rackmount and multi-GPU server environments and uses server-oriented thermal configurations.

For users who want maximum performance in a professional desktop or tower workstation, the RTX PRO 6000 Workstation Edition is the better fit.