NVIDIA GB200 NVL72 – Grace Blackwell Rack-Scale AI Supercomputer
NVIDIA RTX 6000 Ada Generation 48GB – Professional Workstation GPU
NVIDIA RTX PRO 6000 Blackwell 96GB – Professional Workstation GPU
The NVIDIA RTX PRO 6000 Blackwell Workstation Edition is a flagship professional GPU built on the NVIDIA Blackwell architecture for advanced AI, professional visualization, 3D rendering, engineering, simulation, scientific computing, and content creation.
With 24,064 CUDA cores, 96GB of ECC-protected GDDR7 memory, 1,792GB/s memory bandwidth, fifth-generation Tensor Cores, and fourth-generation RT Cores, the RTX PRO 6000 is designed for demanding professional applications requiring exceptional GPU compute performance, large memory capacity, and enterprise-class reliability.
Description
The NVIDIA RTX PRO 6000 Blackwell Workstation Edition is a high-end professional graphics and compute accelerator designed for users working with extremely demanding AI, visualization, simulation, rendering, and engineering workloads.
Powered by the NVIDIA Blackwell architecture, the GPU incorporates 24,064 CUDA cores, fifth-generation Tensor Cores, and fourth-generation Ray Tracing Cores. This combination provides substantial acceleration for traditional GPU computing, AI-assisted workflows, physically based rendering, real-time ray tracing, and professional visualization.
A major advantage of the RTX PRO 6000 is its 96GB of GDDR7 memory with ECC. The large memory capacity allows professionals to work with complex 3D scenes, large AI models, high-resolution datasets, massive CAD assemblies, scientific workloads, and multi-application workflows without being constrained by the memory capacity typically available on consumer graphics cards.
The GPU uses a 512-bit memory interface and provides up to 1,792GB/s of memory bandwidth, allowing large datasets and graphics assets to be transferred rapidly between the GPU and its local memory.
For AI workloads, NVIDIA specifies up to 4,000 AI TOPS using FP4 with sparsity, while single-precision performance reaches approximately 125 TFLOPS and RT Core performance reaches approximately 380 TFLOPS based on NVIDIA’s published specifications.
The RTX PRO 6000 also incorporates four ninth-generation NVENC encoders and four sixth-generation NVDEC decoders, making it suitable for professional video production, real-time video processing, encoding, decoding, and advanced content-creation workflows.
The card supports MIG (Multi-Instance GPU) configurations, allowing its 96GB GPU memory to be partitioned into up to four 24GB instances, two 48GB instances, or one full 96GB instance.
3. Complete Technical Specification
| Specification | Details |
|---|---|
| Manufacturer | NVIDIA |
| Product Family | NVIDIA RTX PRO |
| Model | RTX PRO 6000 Blackwell |
| Variant | Workstation Edition |
| Architecture | NVIDIA Blackwell |
| GPU Architecture Code | GB202 |
| CUDA Cores | 24,064 |
| Tensor Cores | 5th Generation |
| Ray Tracing Cores | 4th Generation |
| AI Performance | Up to 4,000 AI TOPS |
| FP32 Performance | Up to 125 TFLOPS |
| RT Core Performance | Up to 380 TFLOPS |
| GPU Memory | 96GB |
| Memory Type | GDDR7 |
| Memory ECC | Yes |
| Memory Interface | 512-bit |
| Memory Bandwidth | 1,792GB/s |
| System Interface | PCIe 5.0 x16 |
| Display Outputs | 4 × DisplayPort 2.1b |
| Video Encoders | 4 × 9th-Generation NVENC |
| Video Decoders | 4 × 6th-Generation NVDEC |
| MIG Support | Yes |
| MIG Configurations | 4 × 24GB / 2 × 48GB / 1 × 96GB |
| Maximum Power | 600W |
| Power Connector | 1 × PCIe CEM5 16-pin |
| Thermal Design | Double-flow-through |
| Form Factor | 5.4″ H × 12.0″ L |
| Slot Width | Dual-slot |
| Graphics APIs | DirectX 12, Shader Model 6.6, OpenGL 4.6, Vulkan 1.3 |
| Compute APIs | CUDA 12.8, OpenCL 3.0, DirectCompute |
| Primary Market | Professional Workstations |
| Typical Applications | AI, rendering, CAD, simulation, visualization, scientific computing |
Specifications above are based on NVIDIA’s published RTX PRO 6000 Blackwell Workstation Edition documentation.
4. Applications
The NVIDIA RTX PRO 6000 Blackwell is suitable for:
- Artificial intelligence and machine learning
- Generative AI
- AI model development and inference
- Professional 3D rendering
- Real-time ray tracing
- CAD and engineering
- CAE and simulation
- Scientific computing
- Digital twins
- Architecture, engineering and construction
- Product design
- Media and entertainment
- Film and animation production
- Video editing and professional encoding
- Virtual production
- High-resolution visualization
- Virtual reality and extended reality
- Medical visualization
- Geospatial applications
- Data science
- Large-scale datasets and visualization
- Professional workstation applications
The RTX PRO 6000 is specifically positioned by NVIDIA for AI, science, design, rendering, 3D graphics, and video workloads.
5. Compatibility
Hardware Compatibility
The RTX PRO 6000 Blackwell Workstation Edition uses:
- PCIe 5.0 x16
- Dual-slot workstation form factor
- 16-pin PCIe CEM5 power connector
- Double-flow-through cooling
- 600W maximum board power
It should be installed in a workstation chassis with adequate:
- PCIe x16 physical clearance
- Power-delivery capacity
- 16-pin GPU power support
- Airflow
- Thermal capacity
- Motherboard and BIOS compatibility
Software Compatibility
The GPU is designed for professional applications supporting NVIDIA RTX, CUDA, OptiX, Tensor Core, and RTX ray-tracing technologies. NVIDIA provides enterprise/professional drivers and extensive ISV certification support for professional workstation applications.
Important Variant Note
The RTX PRO 6000 Blackwell family has three major configurations:
| Variant | Memory | Power | Cooling / Form Factor |
|---|---|---|---|
| Workstation Edition | 96GB GDDR7 ECC | 600W | Dual-slot, double-flow-through |
| Max-Q Workstation Edition | 96GB GDDR7 ECC | 300W | Dual-slot, active cooling |
| Server Edition | 96GB GDDR7 ECC | 400–600W configurable | Server-oriented passive/ liquid configurations |
NVIDIA lists all three as PCIe Gen 5 x16 products, but their power, thermal and physical configurations differ substantially.
6. Key Benefits
- 96GB ECC GDDR7 memory
- Extremely high 1,792GB/s memory bandwidth
- 24,064 CUDA cores
- Fifth-generation Tensor Cores
- Fourth-generation Ray Tracing Cores
- Up to 4,000 AI TOPS
- Up to 125 TFLOPS FP32
- Approximately 380 TFLOPS RT performance
- PCIe 5.0 x16 interface
- Professional-grade ECC memory
- Advanced hardware ray tracing
- Hardware AI acceleration
- Up to four independent MIG GPU instances
- Four 9th-generation NVENC encoders
- Four 6th-generation NVDEC decoders
- Four DisplayPort 2.1b outputs
- Designed for professional and enterprise applications
- Excellent for large AI models and high-resolution professional datasets
- Enterprise-class professional drivers and ISV support
7. SEO Title
NVIDIA RTX PRO 6000 Blackwell 96GB GDDR7 ECC GPU – PCIe 5.0 Professional AI & Workstation Graphics Card
Related products
-
HPE 881457-B21 2.4TB 10K RPM SAS 12Gb/s SFF Hard Drive
Read more
HPE 881457-B21 2.4TB 10K RPM SAS 12Gb/s SFF Hard Drive
The HPE 881457-B21 is a 2.4TB enterprise-class SAS hard drive designed for mission-critical server environments. Featuring 10,000 RPM performance, 12Gb/s SAS connectivity, 2.5-inch Small Form Factor (SFF), 512e formatting, and an HPE Smart Carrier, it provides a strong combination of storage capacity, performance, and enterprise reliability for compatible HPE ProLiant servers.
Quick View -
Avision AD230 Document Scanner
Read more
Avision AD230 Document Scanner
The Avision AD230 Series is a fast and reliable duplex document scanner designed for efficient business document digitization. With scanning speeds of up to 40 pages per minute, advanced paper-feeding technology, and a reverse roller mechanism, it is well suited for scanning documents, business cards, plastic ID cards, and heavier card stock while maintaining dependable paper handling.
Quick View -
Cisco C9300X-24HX-E Catalyst 9300 24-Port 10G/mGig UPOE+ Switch
Read more
Cisco C9300X-24HX-E Catalyst 9300 Series Switch is a high-performance, stackable enterprise switch featuring 24 x 10G multigigabit copper ports with Cisco UPOE+, supporting speeds from 100 Mbps up to 10 Gbps and delivering up to 90W of power per port. Designed for secure high-speed wired and wireless access, it supports modular uplinks, StackWise-1T, StackPower+, and Cisco IOS XE, making it ideal for demanding enterprise, campus, IoT, and smart-building deployments.
Quick View -
Dell PowerEdge R660
Read more
The Dell PowerEdge R660 is a high-performance 1U dual-socket enterprise rack server designed for high-density virtualization, databases, analytics, mixed workloads, and demanding data-center applications. It supports up to two 4th Generation Intel Xeon Scalable processors with up to 56 cores per processor, 32 DDR5 DIMM slots with up to 8TB RDIMM memory, high-speed SAS/SATA/NVMe storage, PCIe Gen 4 and Gen 5 expansion, optional accelerators, and iDRAC9 remote management. Current Dell documentation also supports configurations with 5th Generation Intel Xeon Scalable processors with up to 64 cores per processor.
Quick View
