Brand: ASUS | Category: Servers
SKU: ESC4000A-E12 | Part #: ESC4000A-E12 | MPN: ESC4000A-E12
Contact for Pricing — Request a Quote
The ASUS ESC4000A-E12 (part number ESC4000A-E12) is a 2U GPU server configured with 4x NVIDIA H100 PCIe 80GB GPUs. It is built on a single-processor AMD EPYC™ 9004/9005 platform and supports four dual-slot GPUs, making it suited to GPU-accelerated compute workloads in data centre and server room environments.
Memory is provided through 12 slots in a 12-channel layout, supporting DDR5 4800/4400 RDIMM and RDIMM 3DS modules up to a maximum of 3TB. Expansion includes 8 PCIe 5.0 slots, storage support for six NVMe drives, and OCP 3.0 networking. Rear I/O comprises 2 x USB 3.2 Gen1 ports, 2 x Gigabit LAN ports, 1 x management port and 1 x VGA port.
For administration, the server includes ASUS Control Center and on-board ASUS ASMB11-iKVM for remote management.
Send the part number through our RFQ form and our team will reply within 48 hours.
| Brand | ASUS |
| Category | Servers |
| SKU | ESC4000A-E12 |
| Part Number | ESC4000A-E12 |
| Condition | New |
| Manufacturer Part Number | ESC4000A-E12 |
| Processor | AMD EPYC™ 9004/9005 single-processor |
| Form Factor | 2U GPU server |
| GPU Support | four dual-slot GPUs |
| Memory Total Slots | 12 (12-channel) |
| Memory Capacity | Maximum up to 3TB RDIMM |
| Memory Type | DDR5 4800/4400 RDIMM/RDIMM 3DS |
| Expansion Slots | 8 PCIe 5.0 slots |
| Storage | six NVMe |
| Networking | OCP 3.0 |
| Rear I/O Ports | 2 x USB 3.2 Gen1 ports, 2 x Gigabit LAN ports, 1 x MAnagement port, 1 x VGA port |
| Management Solution | ASUS Control Center, ASUS ASMB11-iKVM (on-board) |
Reference servers include Dell PowerEdge XE9680 / XE9712, HPE Cray XD670, Lenovo ThinkSystem SR685a / SR675 V3, Supermicro AS-A21GE / SYS-821GE, Gigabyte G593 / G894, ASUS ESC. Share your target platform in the RFQ and we will confirm chassis-to-GPU compatibility and recommended NIC pairing.
Yes — Omnixon supplies the full NVIDIA networking lineup (Quantum-2 / Quantum-X InfiniBand, Spectrum-X Ethernet, ConnectX NICs, BlueField DPUs) so we can quote a complete training-cluster BOM, not just the GPUs.
