Vous recherchez quelque chose ? Assistance technique par chat en direct 24 heures sur 24
Besoin d'aide ? Cliquez ici pour accéder à plus de ressources d'assistance !
Supermicro SYS-422GA-NRT-01-G2 4U Rackmount X14 DP Gold Series GPU SuperServer
- Key Features/Applications: Enterprise AI, Chatbot/RAG, Document Processing, Content Creation, Recommendation, Synthetic Data Generation, Agentic AI, LLM/VLM finetuning/training, HPC, CAD/CAE/CFD, Scientific Research (Modular Dynamics, Weather Forecast, Geological Analysis)
- CPU: Dual Intel® Xeon® 6960P Processor, 72 Cores Per CPU
- Chassis: 4U
- Drive: 1x 960GB M.2 PCIe Gen 4.0 NVMe + 2x 3.8TB U.2 PCIe Gen 4.0 NVMe SSD
- RAM: 16x 64GB DDR5-6400 RDIMM Memory
- GPU: 4x NVIDIA RTX PRO™ 6000 Blackwell Server Edition GPUs
- Network Ports: 2x 10GbE RJ45 LAN Ports (Intel X710-AT2)
- Power Supply: 4x 3200W Redundant Titanium Level Power Supplies
- Pre-Configured Agent Flow AI Inferencing System (PaaS) View Datasheet
* Estimated lead time is currently 2-3 weeks. Actual delivery schedule will depend on quantity and availability.
4U Dual-Processor Gold Series AI Training System
SYS-422GA-NRT-01-G2 Gold Series: A 4U powerhouse with dual 72-core Intel Xeon 6, 1TB DDR5, and 4x NVIDIA Pro 6000 Blackwell GPUs for LLM/VLM fine tuning and Agentic AI.
Unmatched Compute for LLM and VLM Training
This Pre-configured Gold Series 4U rackmount is a titan designed for LLM/VLM finetuning and Generative/Agentic AI. Featuring dual Intel® Xeon® 6960P processors (72 cores per CPU) and 1TB of DDR5-6400 memory, it provides the massive compute density and memory bandwidth required for high-concurrency Chatbot/RAG and Agentic AI deployments.
Elite GPU Acceleration and Versatile Scaling
Pre-installed with four NVIDIA RTX PRO™ 6000 Blackwell Server Edition GPUs, this system is optimized for Physical AI, Scientific Research, and CAD/CAE/CFD. The flexible 4U chassis supports a variety of GPU configurations—up to 8 double-width 600W cards—ensuring your infrastructure can scale alongside evolving Enterprise AI, HPC and Omniverse demands.
Enterprise-Grade Storage and Power Redundancy
Engineered for mission-critical reliability, the system includes high-performance U.2 NVMe SSDs for rapid model loading and four 3200W redundant Titanium power supplies. With integrated 10GbE networking, this Gold Series SKU offers a seamless, ready-to-deploy PaaS foundation for complex geological analysis, weather forecasting, and large-scale document processing.
Key Features
- Fully validated, high-performance, ready-to-deploy AI system
- Dual Intel® Xeon® 6960P Processors for massive parallel processing
- 4x pre-installed NVIDIA RTX PRO™ 6000 Blackwell Server Edition GPUs for next-gen AI efficiency
- High-bandwidth 16-channel 1TB DDR5-6400 RDIMM memory for large dataset hosting
- High-speed U.2 NVMe storage for rapid data caching and model loading
- Quad 3200W Titanium power supplies for 24/7 mission-critical uptime
- 4U chassis supports up to 8 double-width 600W GPUs
- Optimized for Agentic AI: Purpose-built for Agent Flow and advanced PaaS inferencing
| System Family | GPU |
|---|---|
| CPU | Intel® Xeon® 6 Processors with P-cores |
| Specification |
| We're sorry, an error has occurred while generating this content. |