NVIDIA DGX B300 10U Blackwell Ultra system with 8 B300 GPUs, 2.1 TB HBM3e, ConnectX-8 800G networking, BlueField-3 DPUs, and flexible AC or DC data center power options.
Fast Shipping
100k+ SKUs
8/5 Support
Certified
NVIDIA DGX B300 8-GPU Blackwell Ultra AI System | DGX-B300
NVIDIA DGX B300 8-GPU Blackwell Ultra AI System (DGX-B300)
NVIDIA DGX B300 is a Blackwell Ultra DGX system for AI reasoning, training, and high-throughput inference. It integrates eight Blackwell Ultra SXM GPUs with 2.1 TB of aggregate GPU memory, Intel Xeon 6776P host processors, eight ConnectX-8 network interfaces, BlueField-3 DPUs, NVMe storage, and DGX software in a 10U system.
Availability / RFQ: NVIDIA states DGX B300 systems are shipping now. Exact allocation, AC versus DC power implementation, support, quantity, and lead time should still be confirmed for each deployment. 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-B300 |
|---|---|
| Product Type | DGX 8-GPU Blackwell Ultra AI System |
| GPU Configuration | 8x NVIDIA Blackwell Ultra SXM GPUs |
| GPU Memory | 2.1 TB total |
| CPU | Intel Xeon 6776P processors |
| NVLink Bandwidth | 14.4 TB/s aggregate |
| Networking | 8x ConnectX-8 VPI up to 800 Gb/s; 2x BlueField-3 DPUs up to 400 Gb/s |
| Storage | 2x 1.9 TB M.2 OS drives; 8x 3.84 TB E1.S NVMe internal |
| Form Factor / Power | 10U / approximately 14 kW |
| Availability Status | Available / RFQ |
Deployment and Compatibility Considerations
- Select AC/PDU or DC/busbar power implementation according to the target rack standard.
- Plan 800G scale-out networking, optics, cables, and switch radix around the eight ConnectX-8 interfaces.
- Verify rack cooling, power redundancy, service access, and facility readiness.
- Size external AI storage and cluster management for reasoning, training, and inference workloads.
Common Use Cases
- AI reasoning and test-time scaling
- Large-model training and post-training
- High-throughput inference
- DGX SuperPOD and AI factory deployments
Frequently Paired Infrastructure
- ConnectX-8 800G fabrics
- BlueField-3 DPUs
- Spectrum-X800 or Quantum-X800 switches
- AI storage and NVMe tiers
- Rack power and cooling infrastructure
Official NVIDIA Documentation
Frequently Asked Questions
Is DGX B300 shipping?
Yes, Contact Catalyst Data Solutions to discuss current lead time, compatibility, configuration requirements, and sourcing options.
How much GPU memory is in DGX B300?
The DGX B300 has 2.1 TB of total GPU memory across eight Blackwell Ultra SXM GPUs.
| product type | GPU Server / DGX AI System |
|---|---|
| Platform / Architecture | NVIDIA DGX B300 / Blackwell Ultra |
| Accelerator Configuration | 8 Blackwell Ultra GPUs / 2.1 TB |
| Networking | 8x ConnectX-8 up to 800 Gb/s |
| Availability Status | Available / RFQ |
