NVIDIABlackwell

RTX PRO 6000 Blackwell

The most powerful desktop GPU ever created. 96GB GDDR7 ECC memory with 1792 GB/s bandwidth, 5th-gen Tensor Cores delivering 5X faster AI inference, and 4th-gen RT Cores with RTX Mega Geometry for up to 100X more ray-traced triangles. Engineered for AI, design, simulation, and content creation professionals.

PRO
6000
Blackwell
96GB GDDR7
96GB GDDR7 ECC Memory
24,064 CUDA Cores
5X Faster AI Inference
600W TDP

Blackwell Professional Features

Next-generation professional capabilities for demanding workloads

5X
Faster AI Inference
FP4 precision with 5th-gen Tensor Cores
100X
More Triangles
RTX Mega Geometry ray tracing
4
MIG Instances
Multi-workload partitioning
4:2:2
Video Support
9th-gen NVENC/NVDEC

RTX PRO 6000 Editions

Choose the right edition for your deployment scenario

MAX-Q RTX PRO 6000
Memory 96GB GDDR7 ECC
Bandwidth 1792 GB/s
CUDA Cores 24,064
TDP Optimized
Enterprise Price Contact Sales
Mobile workstations · Power optimized
SERVER RTX PRO 6000
Memory 96GB GDDR7 ECC
Bandwidth 1792 GB/s
CUDA Cores 24,064
Deployment Data Center
Enterprise Price $15,000 - $20,000
Cloud · VDI · Multi-tenant

Technical Specifications

Complete RTX PRO 6000 Blackwell technical details

Specification RTX PRO 6000 Blackwell
Architecture
GPU Architecture NVIDIA Blackwell
CUDA Cores 24,064
Tensor Cores 5th Generation
RT Cores 4th Generation
Memory
Memory Capacity 96GB GDDR7
Memory Bandwidth 1792 GB/s
Memory Interface 512-bit
ECC Memory Yes
Performance
AI Inference (vs prev gen) Up to 5X faster (FP4)
Ray Tracing 2X intersection rate + RTX Mega Geometry
Triangle Capacity Up to 100X more ray-traced triangles
Media Engine
NVENC 4x 9th Generation (4:2:2 H.264/HEVC)
NVDEC 4x 6th Generation (4:2:2 decode)
AV1 Support Encode and Decode
Connectivity
PCIe PCIe 5.0 x16
Display Outputs DisplayPort 2.1b
Form Factor Dual-slot PCIe
Power & Thermal
TDP (Workstation) 600W
Cooling Double-flow-through design
Enterprise Features
MIG Support Up to 4 instances
vGPU Support Yes
RTX Studio Certified Yes
Enterprise Drivers Yes

Professional Use Cases

RTX PRO 6000 Blackwell powers demanding professional workflows

Agentic AI Development

Build and deploy AI agents locally with 5X faster inference. 96GB memory supports large language models and complex multi-agent systems.

3D Design & CAD

RTX Mega Geometry handles 100X more triangles for complex architectural, automotive, and industrial design visualization.

Content Creation

9th-gen NVENC with 4:2:2 support accelerates professional video workflows. RTX Studio certified for creative applications.

Scientific Simulation

ECC memory ensures data integrity for computational fluid dynamics, structural analysis, and physics simulations.

Virtual Workstations

Server Edition with MIG partitioning enables multi-user VDI deployments. Securely share GPU resources across teams.

Medical Imaging

Accelerate medical image reconstruction, AI-assisted diagnostics, and 3D visualization of patient data.

Professional GPU Comparison

How RTX PRO 6000 compares to previous professional GPUs

RTX PRO 6000 vs RTX 6000 Ada

  • Architecture: Blackwell vs Ada Lovelace
  • Memory: 96GB GDDR7 vs 48GB GDDR6
  • AI Inference: 5X faster (FP4)
  • Ray Tracing: 2X intersection + Mega Geometry
  • Video: 4:2:2 vs 4:2:0
  • TDP: 600W vs 300W
Verdict: RTX PRO 6000 for AI-intensive and complex geometry workloads

RTX PRO 6000 vs Data Center GPUs

  • Memory Type: GDDR7 vs HBM3e
  • Memory Size: 96GB vs 192-288GB
  • Ray Tracing: Yes vs Limited/No
  • Video Encode: Advanced vs Basic
  • Target: Professional/Workstation vs AI Training
  • Deployment: Desktop/Server vs Data Center
Verdict: RTX PRO 6000 for visualization + AI; Data Center GPUs for pure AI training

Frequently Asked Questions

What is the NVIDIA RTX PRO 6000 Blackwell?

The RTX PRO 6000 Blackwell is NVIDIA's most powerful professional desktop GPU, featuring 96GB GDDR7 ECC memory, 24,064 CUDA cores, 5th-gen Tensor Cores, and 4th-gen RT Cores. It delivers 5X faster AI inference than previous generation and supports advanced features like RTX Mega Geometry for professional workflows.

What editions of RTX PRO 6000 are available?

RTX PRO 6000 is available in three editions: Workstation Edition (600W, full desktop performance), Max-Q Workstation Edition (optimized for mobile workstations with reduced power), and Server Edition (for cloud and data center deployment with enhanced virtualization and MIG support).

What is RTX Mega Geometry?

RTX Mega Geometry is a new feature in 4th-gen RT Cores that enables up to 100X more ray-traced triangles by efficiently handling complex geometric detail. This is essential for architectural visualization, automotive design, and other applications with highly detailed 3D models.

Does RTX PRO 6000 support MIG?

Yes, RTX PRO 6000 supports MIG (Multi-Instance GPU) with up to 4 partitions, allowing secure partitioning of a single GPU for multi-user or multi-workload deployments. This is particularly useful for VDI and cloud workstation scenarios.

What cooling does RTX PRO 6000 require?

The RTX PRO 6000 Workstation Edition features a double-flow-through cooling design optimized for sustained 600W operation in professional workstations. The Server Edition is designed for data center cooling infrastructure.

Is RTX PRO 6000 suitable for AI development?

Yes, RTX PRO 6000 is excellent for AI development with 96GB memory supporting large models, 5X faster inference with FP4 precision, and professional-grade reliability. It's ideal for building AI agents, fine-tuning models, and running inference locally.

Deploy NVIDIA RTX PRO 6000

Get enterprise RTX PRO 6000 workstations and servers through our OEM partnerships with expert configuration support.

Available through:
Dell HPE Lenovo Cisco Supermicro
Reconnecting to the server...

Please wait while we restore your connection

An unhandled error has occurred. Reload 🗙