NVIDIA DGX GB300 NVL72 rack-scale Blackwell Ultra system with 72 GPUs, 36 Grace CPUs, about 20 TB GPU memory, NVLink, and DGX software.
Fast Shipping
100k+ SKUs
8/5 Support
Certified
NVIDIA DGX GB300 NVL72 | Rack-Scale Blackwell Ultra AI System
NVIDIA DGX GB300 NVL72 Rack-Scale AI System (DGX-GB300-NVL72)
NVIDIA DGX GB300 NVL72 is the DGX rack-scale implementation of Blackwell Ultra NVL72 for AI reasoning, test-time scaling, training, and high-throughput inference. It combines 72 Blackwell Ultra GPUs and 36 Grace CPUs in a liquid-cooled NVLink domain with NVIDIA DGX system software and lifecycle-managed firmware and drivers.
Availability / RFQ: This product is a current NVIDIA platform or component and is listed for RFQ. Pricing, allocation, quantity, configuration, support, and lead time can vary with current channel conditions. Contact Catalyst Data Solutions to discuss current status, compatibility, configuration requirements, and sourcing options.
End-user / export compliance notice: Advanced AI accelerators and systems may be subject to applicable U.S. export controls, end-user/end-use restrictions, destination requirements, and OEM/channel policies. Catalyst may request purchaser, final end-user, intended-use, and deployment-location information before a quote or fulfillment can be approved. An RFQ does not guarantee allocation or shipment.
Technical Specifications
| SKU | DGX-GB300-NVL72 |
|---|---|
| Product Type | DGX Blackwell Ultra Rack-Scale AI System |
| GPU / CPU Configuration | 72 Blackwell Ultra GPUs / 36 Grace CPUs |
| GPU Memory | Approximately 20 TB HBM3e aggregate |
| GPU Memory Bandwidth | Up to 576 TB/s aggregate |
| NVLink Bandwidth | 130 TB/s |
| CPU Memory | 17 TB LPDDR5X aggregate |
| Cooling | Fully liquid-cooled rack-scale architecture |
| Software | NVIDIA DGX system software stack and active release notes |
| Availability Status | Available / RFQ |
Deployment and Compatibility Considerations
- Treat GB300 NVL72 as a rack-scale facility project with liquid cooling, high-density power, network, and storage design.
- Validate exact DGX release compatibility, firmware, Mission Control/management, support, and cluster software.
- Plan Quantum-X800 or Spectrum-X scale-out network, ConnectX-8 connectivity, optics, and cable maps.
- Size storage around long-context inference, checkpointing, and AI reasoning workloads.
Common Use Cases
- AI reasoning and test-time scaling
- Massive agentic inference
- Large-model training
- DGX SuperPOD and AI factory deployments
Frequently Paired Infrastructure
- ConnectX-8 and BlueField infrastructure
- Quantum-X800 or Spectrum-X800
- AI-native storage
- High-density rack power and liquid cooling
Official NVIDIA Documentation
NVIDIA DGX GB300 NVL72 Release Notes
Frequently Asked Questions
What generation of GPU does DGX GB300 NVL72 use?
It uses NVIDIA Blackwell Ultra GPUs.
Is DGX GB300 software actively maintained?
Yes. NVIDIA publishes active DGX GB300 NVL72 release notes and system-software package information.
| product type | DGX Rack-Scale AI System |
|---|---|
| Platform / Architecture | NVIDIA Blackwell Ultra GB300 NVL72 |
| Accelerator Configuration | 72 Blackwell Ultra GPUs / 36 Grace CPUs |
| Cooling Type | Liquid-Cooled Rack-Scale |
| Availability Status | Available / RFQ |
