| Customization: | Available |
|---|---|
| Application: | Enterprise Level |
| Instruction System: | CISC |
| Shipping Cost: | Contact the supplier about freight and estimated delivery time. |
|---|
| Payment Methods: |
|
|---|---|
| Support payments in USD |
| Secure payments: | Every payment you make on Made-in-China.com is protected by the platform. |
|---|
| Refund policy: | Claim a refund if your order doesn't ship, is missing, or arrives with product issues. |
|---|
Suppliers with verified business licenses
Audited Supplier Audited by an independent third-party inspection agency
INSPUR NF5688M7 GPU_Nvidia_HGX-H20-8GPU_768G INTEL8480*2 6U RACK Server for Ai Training
Wave NF5688M7
6U rackmount server CPU: 2 CPU_Intel_8480+_Xeon_2.0GHz_56C_105M_350W; Memory: 2048GB (Mem_64G_DDR5-4800MHz*32), supports up to 32 memory slots; Hard disk: 4 pieces of 7.68TB SATA SSD System disk: 2 pieces of 1.92TB SATA SSD NIC: 4-port Gigabit network interface, 1 dual-port 10GbE NIC with optical module, 2 HCA 2Port HDR200G; GPU card: GPU-Nvidia-H20-8GPURAID card: support RAID 0/1/5/6 Cache 4GB with supercapacitor; Power supply: PowerSupply_ 3200W_Titanium_220VACor240VDC*2, PowerSupply_2700W_Titanium_220VACor240VDC*6, with power cord; standard rail 3 years manufacturer's original after-sales service, 7*24 response
| model | specific |
| NF5688M7 | CPU:2*intel 8480+56C 2.0G Memory: 32*64G DDR5 System Disk: 2*960G SATA SSD Cache Disk: 4*3.84T NVME Raid Card: 1*Raid Card NIC: 1*Dual Port 25G NIC + Module HCA Card: 4*NDR Single Port 400G + 1*Dual Port HDR200G Power supply: redundant power supply Warranty: Three-year warranty |
| 33 | Intelligent computing center project full ecological support | AI software | Large model training push cloud platform-manage all variety of large models | Price negotiable! | One-stop model service: to assist customers to complete the private domain large model development and hosting | |||
| 34 | IB networking | The 128-node networking | Package network installation, debugging, operation and maintenance | Price negotiable! | Containing HPC cluster, GPU / NPU cluster software, provide compatible accessories | |||
| 35 | memory | Large model high price distributed storage, and NV cooperation experiment number | ¥175 W | OR Price negotiable! | Support 200G400G, full flash file storage XNVIDIA InfniBand | |||
| According to (128 computing nodes, read: 500 GB / S write: 250 GB / S), the price and performance are better than a domestic server factory | Provide efficient, flexible and economical solutions for HPC / Al computing (support | |||||||
| Holding Odm, Oem, white card server, etc.) | ||||||||
| 36 | rent | H100 Nvlink | 8468*2/3.84T * 2640gb 8 card | ¥7.8 W | 128 units: 5 mb bandwidth, station / month (5 years) | Pay three three, you need to check the lessee's ability to pay | ||
| 37 | ¥8.3 W | 128 units: 5 mb bandwidth, station / month (3 years) | ||||||
| 38 | RTX4090*8 | 6133 *2/64G/2T M.2 4U8 card | ¥0.55 W | 50 units: 5 mb bandwidth, station / month (5 years) | ||||
| 39 | pressure measurement | GPU module and each brand of Nvlink complete machine, inspection and tuning ("disassembly and replacement" | About ¥0.3W | The service contents include: 1, the machine lights up the machine; 2, replacement system; 3, pressure measurement through the test workshop of dust-free, silent, full power supply, and 60 sets | ||||
| serve) | Nvlink Machine, and then directly to the machine room, on the shelves carefree. | |||||||
| 40 | maintenance | The maintenance of intelligent computing center includes two types of warranty and warranty | Price negotiable! | The service content is the service corresponding time of the nose part; the module generation warranty, etc | ||||
){kind=link}