
ASUS XA NB3I-E12
9U NVIDIA HGX B300 8-GPU server with dual Intel Xeon 6 for heavy AI training and inference.
- GPUs: NVIDIA HGX B300 8-GPU
- CPUs: Dual Intel Xeon 6
- Networking: 8× embedded ConnectX-8 InfiniBand, dual 10 GbE
Model ESC8000A-E13X
4U NVIDIA RTX PRO Server with ConnectX-8 SuperNICs and eight RTX PRO 6000 Blackwell GPUs.
The ASUS ESC8000A-E13X is a 4U NVIDIA RTX PRO Server based on dual AMD EPYC 9005 processors. It holds eight RTX PRO 6000 Blackwell Server Edition GPUs connected through ConnectX-8 SuperNICs with eight 400 Gb/s ports.
Built on the NVIDIA MGX architecture with PCIe Gen6 GPU slots and a tool-less design, it is aimed at enterprise AI factories, agentic AI and industrial simulation.
Dual-slot GPUs, up to 600 W each.
Eight 400 Gb/s QSFP ports for GPU-to-GPU traffic.
Up to 192 Zen 5c cores per socket, 12-channel DDR5-6400.
Eight Gen6 x16 GPU slots plus Gen5 slots for NICs and DPUs.
Modular architecture for fast, large-scale deployment.
Quick maintenance and upgrades.
| CPUs | 2× Socket SP5, AMD EPYC 9005, up to 500 W cTDP |
|---|---|
| Memory slots | 24 (12 channels per CPU) |
| Memory capacity | Up to 3 TB DDR5-6400 / 5600 RDIMM |
| GPUs | Up to 8× NVIDIA RTX PRO 6000 Blackwell Server Edition |
|---|---|
| Rear slots | 8× PCIe x16 Gen6 (dual-slot GPU, FH/FL); 1× PCIe Gen5 x16 (NIC / BlueField-3) |
| Front slot | 1× PCIe x16 (Gen5 x8, FH/HL) |
| Storage | 8× 2.5" front hot-swap; 2× M.2 (PCIe Gen5 x4, up to 22110) |
|---|---|
| Networking | 8× 400 Gb/s QSFP (ConnectX-8); 2× 10 GbE (X710-AT2); 1× management |
| Form factor | 4U |
|---|---|
| Dimensions | 800 × 439.5 × 175 mm |
| Weight | 45 kg net / 56 kg gross |
| Power | 3+1 redundant 3200 W 80 PLUS Titanium |
| Operating temperature | 10 °C to 35 °C |
| Compliance | BSMI, CB, CE, FCC (Class A), RCM |
Tell us your workload and timeline for the ASUS ESC8000A-E13X. We reply within one business day with a configuration, pricing and lead time.