Product
News
- actively responds to the environmental protection call and launches a new series
- More and more small and medium-sized enterprises choose Wundorful International
- In response to the unique needs of different customers
- In order to further enhance the user experience,
- announced today the launch of a new series of cross-border servers
- Holds an Online Server Technology Exchange Meeting to Promote
Contact us
E-mail:[email protected]
Address:RM 1002,10/F,EASEY COMM BUILDING,253-261 HENNESSY ROAD ,WAN CHAI,HONG KONG
HPC Server
AIScale AI-10000 Deep Learning Workstation
- Model:AIScale AI-10000
- CPU:AMD Ryzen Threadripper PRO 7995WX (96C/192T, 4.0GHz Base, 384MB L3 Cache)
- MotherBoard:ASUS Pro WS TRX50-SAGE WIFI (8-channel DDR5, 7×PCIe 5.0 slots)
- Memory:SK Hynix 256GB DDR5 ECC RDIMM (64GB×4, 5200MHz)
- Storage:Samsung 990 PRO 2TB NVMe+WD Ultrastar DC HC560 20TB HDD ×4
- Chassis:Phanteks Enthoo Elite (Full-Tower, dual-system support)
- Power:Seasonic PRIME TX-1600W (1600W, 80Plus Titanium)
- GraphicsCard:NVIDIA RTX 6000 Ada Generation 48GB GDDR6 ×2 (with NVLink)
- NetworkCard:Dual-port Intel X550-T2 10GbE + WiFi 6E
- Price:$12999
Model: AIScale AI-10000 Deep Learning Workstation
Key Specifications:
- CPU: AMD Ryzen Threadripper PRO 7995WX (96C/192T, 4.0GHz Base, 384MB L3 Cache)
- Cooling: Noctua NH-U14S TR5-SP6 ×2 (Threadripper PRO optimized)
- Motherboard: ASUS Pro WS TRX50-SAGE WIFI (8-channel DDR5, 7×PCIe 5.0 slots)
- Memory: SK Hynix 256GB DDR5 ECC RDIMM (64GB×4, 5200MHz)
- Storage:
- OS Drive: Samsung 990 PRO 2TB NVMe
- Data Drives: WD Ultrastar DC HC560 20TB HDD ×4
- Cache: Intel Optane P5800X 800GB ×2
- GPU: NVIDIA RTX 6000 Ada Generation 48GB GDDR6 ×2 (with NVLink)
- PSU: Seasonic PRIME TX-1600W (1600W, 80Plus Titanium)
- Expansion: Mellanox ConnectX-6 DX 100GbE Dual-port
- Chassis: Phanteks Enthoo Elite (Full-Tower, dual-system support)
- OS: Ubuntu 22.04 LTS / Windows 11 Pro for Workstations
Professional Features:
- Extreme Multicore: 96-core Threadripper PRO for parallel workloads
- Dual-GPU NVLink: 96GB effective memory for large models
- Hybrid Storage: Optane cache + HDD capacity
- High-Speed Networking: 100GbE with RDMA support
- Advanced Cooling: Dual-tower heatsinks + optimized airflow
Use Cases:
- Large-scale AI model training
- Scientific computing & simulation
- 8K/12K real-time rendering
- Autonomous driving development
- Financial risk modeling