ขอใบเสนอราคา (Request Quote) — ทักไลน์ @gearlab หรือ https://m.me/gearlabth
Supermicro SYS-822GS-NB3RT 8U ระบบ AI Server ระดับ Enterprise บน NVIDIA HGX B300 NVL8 8-GPU Platform รุ่น Full InfiniBand Fabric เหมาะสำหรับการขยายระบบเป็น Multi-Node Cluster ขนาดใหญ่
| Form Factor |
8U Rackmount |
| GPU |
NVIDIA HGX B300 NVL8, 8x B300 GPU, Air-Cooled (288GB HBM3e รวม) |
| CPU |
2x Intel Xeon 6768P |
| RAM |
32x 96GB DDR5-6400 ECC RDIMM (3TB รวม) |
| Storage |
2x 960GB M.2 NVMe PCIe 4.0 SSD, 2x 7.68TB E1.S NVMe PCIe Gen5 SSD |
| Networking (Compute Fabric) |
8x NVIDIA ConnectX-8 SuperNIC 800G OSFP InfiniBand (หนึ่งต่อ GPU) — รองรับ Multi-Node Scale-Out |
| Networking (Management) |
1x NVIDIA ConnectX-6 Lx Dual-Port 25GbE SFP28, 4x 25GbE SFP28 Optical Transceivers |
| Power |
6x 6600W Titanium Redundant PSU (3+3) |
| Security / Management |
TPM 2.0, Tool-less Rail Kit, SDDC Remote Management License |
| NIC Reference |
MELLANOX-MCX755106AS-HEAT |
การใช้งานที่เหมาะสม: Distributed Multi-Node AI Training Cluster, Large-Scale LLM Training, Generative AI ระดับ Data Center, HPC ขนาดใหญ่
Key Applications
- Scientific Research
- Conversational AI
- Business Intelligence & Analytics
- Drug Discovery
- Finance Services and Fraud Detection
- AI/Deep Learning Training and Inference
- Large Language Model (LLM) and Generative AI
- High Performance Computing (HPC)
- Autonomous Vehicle Technologies
Key Features
- Dual Intel® Xeon® 6700 series processors with P-cores, up to 350W TDP
- NVIDIA HGX B300 8-GPU with 5th Generation NVLink® 1.8TB/s, 2.3TB of HBM3e GPU memory per system
- 8x NVIDIA ConnectX®-8 SuperNICs, up to 800Gb/s, up to two North/South NICs
- 8x Hot-swap E1.S NVMe storage drive bays and 2x M.2 NVMe boot drives
- 32x DIMMs with up to 8TB at 5200MT/s or up to 4TB at 6400MT/s DDR5 RDIMM
- 6x redundant (3+3) 6.6kW Titanium level power supplies
|