Supermicro SYS-822GS-NB3RT-01-G2 AI GPU Server | Dual Intel Xeon 6768P | 8x NVIDIA HGX B300 GPUs | 2TB DDR5 | 800GbE Networking
Supermicro SYS-822GS-NB3RT-01-G2 AI GPU Server
8x NVIDIA HGX B300 GPUs with Dual Intel Xeon 6768P Processors for Large-Scale AI Training and HPC
Note: New Supermicro product with a 3-year warranty. Available exclusively for U.S. end users. Limited quantities are available for immediate shipment.
The Supermicro SYS-822GS-NB3RT-01-G2 is an ultra-high-performance AI and HPC server engineered for next-generation artificial intelligence, machine learning, and large-scale data analytics workloads. Built on the advanced NVIDIA HGX B300 platform, this enterprise-grade system combines 8 NVIDIA HGX B300 GPUs, dual Intel Xeon 6768P processors, and 2TB DDR5-6400 memory to deliver exceptional compute density and accelerated AI performance.
Designed for training large language models (LLMs), generative AI applications, scientific simulations, and deep learning frameworks, the server provides a total of 256 CPU cores through dual Intel Xeon 6768P processors operating at 2.4GHz. The integrated NVIDIA Blackwell-based HGX B300 GPU architecture enables breakthrough performance for AI model training, inference, and HPC environments.
To support large-scale AI clusters, the SYS-822GS-NB3RT-01-G2 features an 8-port 800GbE/XDR800 networking fabric, ensuring ultra-low latency communication and high-bandwidth data transfer across distributed GPU infrastructures. A 960GB M.2 NVMe SSD provides fast boot and operating system storage, while 2TB of ECC DDR5-6400 memory delivers the bandwidth required for memory-intensive AI workloads.
Built for continuous enterprise operation, the server includes redundant 6600W Titanium-level power supplies, advanced cooling architecture, and enterprise-class reliability features. The platform supports leading AI and HPC software frameworks including NVIDIA CUDA, TensorRT, NCCL, RAPIDS, and NVIDIA AI Enterprise, making it ideal for cloud providers, research institutions, AI factories, and large-scale data centers.
Product Specifications – Supermicro SYS-822GS-NB3RT-01-G2 AI GPU Server
CLICK HERE FOR DATA SHEET
| Specification Category | Details |
|---|---|
| Product Name | Supermicro SYS-822GS-NB3RT-01-G2 |
| Manufacturer | Supermicro |
| Product Type | AI Training & High-Performance Computing (HPC) GPU Server |
| Server Form Factor | 8U Rackmount GPU Server |
| System Architecture | NVIDIA HGX B300 Platform |
| Processor Configuration | Dual-Socket Server |
| CPU Model | 2 × Intel® Xeon® 6768P Processors |
| CPU Cores | 128 Cores per CPU (256 Total Cores) |
| CPU Base Frequency | 2.4 GHz |
| Processor Architecture | Intel Xeon 6 Performance-Core (P-Core) Architecture |
| GPU Configuration | 8 × NVIDIA HGX™ B300 GPUs |
| GPU Architecture | NVIDIA Blackwell Architecture |
| GPU Interconnect | NVIDIA NVLink High-Speed GPU Fabric |
| Target Workloads | AI Training, Generative AI, LLM Training, Deep Learning, HPC, Scientific Computing |
| System Memory | 2 TB DDR5-6400 ECC Registered Memory |
| Memory Technology | DDR5 RDIMM |
| Memory Speed | 6400 MT/s |
| Memory Protection | ECC (Error-Correcting Code) |
| Boot Drive | 1 × 960GB M.2 NVMe SSD |
| Storage Interface | PCIe NVMe |
| Network Connectivity | 8-Port 800GbE / XDR800 High-Speed Fabric |
| Network Speed | Up to 800 Gbps per port |
| Networking Technology | High-Speed Ethernet / AI Cluster Interconnect |
| Cluster Scalability | Optimized for Multi-Node AI Superclusters |
| Power Supply Configuration | Redundant Power Supplies |
| Power Rating | 6600W Titanium-Level Efficiency |
| Power Redundancy | Yes |
| Efficiency Certification | 80 PLUS Titanium Equivalent |
| Cooling System | High-Performance Redundant Cooling Architecture |
| Management | Remote Server Management & Monitoring Support |
| Security Features | Secure Boot, TPM Support, Hardware Root of Trust (platform dependent) |
| Expansion Architecture | HGX GPU Baseboard with High-Speed PCIe Infrastructure |
| AI Framework Support | NVIDIA CUDA, TensorRT, NCCL, RAPIDS, AI Enterprise |
| Operating System Support | Linux Enterprise Distributions, AI/HPC Software Stacks |
| Deployment Environment | AI Data Centers, Research Labs, Cloud AI Infrastructure |
| Use Cases | Large Language Models (LLMs), AI Training, AI Inference, Scientific Simulation, Data Analytics |
| Reliability Features | Redundant Power, ECC Memory, Enterprise-Class Components |
FAQs
1. How many GPUs does the SYS-822GS-NB3RT-01-G2 support?
It comes with 8 NVIDIA HGX B300 GPUs for advanced AI and HPC workloads.
2. Which processors are installed in this server?
The system features 2× Intel Xeon 6768P processors with a total of 256 CPU cores.
3. How much memory is included?
It is equipped with 2TB DDR5-6400 ECC memory for memory-intensive applications.
4. What networking does the server offer?
The server includes 8-port 800GbE/XDR800 networking for high-speed cluster connectivity.
5. What workloads is this server designed for?
It is optimized for AI training, LLMs, generative AI, machine learning, and HPC workloads.
