NVIDIA B300 (Blackwell Ultra) SXM with 288 GB HBM3e, up to 8 TB/s memory bandwidth, and fifth-generation NVLink.
Ground, Overnight
Fast Shipping
In Stock
100k+ SKUs
Sales and Support
8/5 Support
New / Refurbished
Certified
NVIDIA B300 SXM Blackwell Ultra Data Center GPU | B300-SXM
⭐ Top Rated Verified Buyers
We are trusted for quality, reliability and fast shipping 💯
0
People watching this product now!
Description
NVIDIA B300 SXM Blackwell Ultra Data Center GPU (B300-SXM)
NVIDIA B300 SXM is the Blackwell Ultra data center GPU used in HGX B300, DGX B300, and related OEM systems. It expands memory capacity to 288 GB of HBM3e per GPU while retaining the high-bandwidth SXM and NVLink architecture required for dense multi-GPU AI systems.
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 | B300-SXM |
|---|---|
| Product Type | Blackwell Ultra AI / HPC GPU |
| Architecture | NVIDIA Blackwell Ultra |
| Form Factor | SXM platform GPU |
| GPU Memory | 288 GB HBM3e |
| Memory Bandwidth | Up to 8 TB/s per GPU |
| GPU Interconnect | Fifth-generation NVIDIA NVLink |
| NVLink Bandwidth | Up to 1.8 TB/s GPU-to-GPU |
| Common Platforms | HGX B300, DGX B300, OEM Blackwell Ultra systems |
| Availability Status | Available / RFQ |
Deployment and Compatibility Considerations
- Deploy only in validated Blackwell Ultra SXM/HGX or OEM systems.
- Confirm OEM baseboard, firmware, host CPU, system memory, power delivery, cooling, and service requirements.
- Plan 800G-class scale-out networking where required by the selected server and cluster architecture.
- Size storage and data pipelines for reasoning, long-context inference, training checkpoints, and model datasets.
Common Use Cases
- AI reasoning and test-time scaling inference
- Large-model training and post-training
- Agentic AI inference
- High-density enterprise and cloud AI systems
Frequently Paired Infrastructure
- NVIDIA HGX B300 or DGX B300
- ConnectX-8 SuperNICs
- BlueField-3 DPUs
- Spectrum-X800 or Quantum-X800 fabrics
- High-throughput AI storage
Official NVIDIA Documentation
NVIDIA HGX Blackwell Ultra Datasheet
Frequently Asked Questions
How much memory does a B300 SXM GPU provide?
NVIDIA documents 288 GB of HBM3e memory per Blackwell Ultra B300 GPU.
Is B300 a drop-in replacement for B200 in any server?
No. Even though both use SXM-class platforms, server support, firmware, cooling, power, baseboard design, and OEM qualification must be verified for B300.
Additional information
| product type | Data Center AI / HPC GPU |
|---|---|
| Architecture / Series | NVIDIA Blackwell Ultra B300 |
| GPU Memory | 288 GB HBM3e |
| Form Factor / Interconnect | SXM / NVLink 5 |
| Availability Status | Available / RFQ |
About brand
NVIDIA is the leader in GPU computing and AI infrastructure for high-performance workloads.
