AD-ASMB-831
AMD® EPYC™ 9004 ATX Server Board with 6 x DDR5, 5 x PCIe x16 + 2 x PCIe x8, , Dual 10GbE, and IPMI

- LGA 6096 (Socket SP5) ATX server board with AMD® EPYC™ 9004 processor
- DDR5 4800 MHz RDIMM up to 384GB (6 DIMMs)
- Five PCIe x16 (Gen5) and two PCIe x8 (Gen5) for four double-deck cards
- Intel® X710-AT2 dual 10GbE ports
- Nine SATA3 and seven USB 3.2 (Gen1) ports
- Two M.2 2280/22110 (SATA / PCIe compatible)
- 0 ~ 60 °C/ 32 ~ 140 °F ambient operating temperature range


Model | ASMB-831T2-00A1 |
Processor System | |
CPU | AMD EPYC™ 9004 processors |
CPU TDP | Up to 210W |
Frequency | Up to 3.0 GHz |
Core Number | Up to 32 cores |
L3 Cache | Up to 256MB |
BIOS | AMI SPI 256 Mbit |
Integrated Graphics | - |
Form Factor | |
Form Factor | ATX |
Memory | |
Capacity | Up to 384 GB |
Technology | DDR5 ECC-REG 4000/4400/4800 MHz |
Socket | 8 x 288-pin DIMM |
Type | RDIMM/ LRDIMM |
ECC Support | Yes |
REG Support | YES |
Display | |
Controller | ASPEED AST2600 |
Graphic Memory | Onboard memory 64MB |
Multiple Display | - |
Display Interface | VGA: Up to 1920 x 1200 |
Expansion Interface | |
PCI Express x 16 | 5 (Gen5) |
PCI Express x 8 | 2 (Gen5) |
Ethernet | |
Controller | LAN1: Intel X710-AT2, LAN2: Intel Intel X710-AT2, LAN3: Realtek 8211F |
Speed | 10/100/1000 Mbps & 10G Base-T |
Connector | 3 x RJ45 |
Watchdog Timer | |
Output | System Reset |
Interval | Programmable 1 ~ 255 sec/min |
I/O | |
SATA | 9 x SATA3 |
SAS Port | - |
SAS controller | - |
USB2.0 | 7 (1 TypeA) |
USB3.0 | 7 |
GPIO | 8-bit |
LPT | - |
COM Port | 1 (1 x RS232) |
PS/2 Keyboard/Mouse | - |
SPDIF | - |
System Fan | 5 |
Rear I/O Interface | |
Display | VGA x1 |
LAN | 3 (One is for remote IPMI/iKVM consloe Only and doesn't have LAN function) |
USB3.0 | 4 |
USB2.0 | - |
COM | 1 (1 x RS232) |
PS/2 | - |
VGA | 1 |
DVI | - |
Power Requirement | |
Current | 2.96A/3.26A/0.61A/15.72A/6.87A |
Voltage | +12V/+5V/+3.3V/+5Vsb/-12V/12V_8P |
Environment | |
Operating-Temperature | 0 ~ 60° C (32 ~ 140° F) |
Non-Operational Temperature | -40 ~ 85° C (-40 ~ 185° F) |
Non-Operating Vibration | 3.5 Grms |
Physical Characteristics | |
Dimensions (LxW) | 304.8 x 244 mm (12" x 9.6") |
Weight | 1.3 kg |
Certification | |
EMC | CE/FCC |
Management Solution | |
Software | IPMI 2.0 |
Out of band remote management | IPMI 2.0 |
ASMB-831 is a single socket server motherboard ready for AMD EPYC™ 9004 series processors and is designed to offer high performance flexible configuration, and high reliability. ASMB-831 Server Motherboard can effectively meet the demands of factory automation, Industry 4.0/5.0, smart city initiatives, and emerging Al applications
Case Study: Powering Autonomous Mining with Rugged Edge AI
Background
The global mining industry is undergoing a significant digital transformation, driven by the adoption of autonomous vehicles, artificial intelligence, and edge computing to enhance safety, efficiency, and sustainability. Harsh and unpredictable environments such as open‑pit mines demand rugged computing solutions capable of real‑time AI inference for obstacle detection, navigation, and predictive maintenance. .
Traditional IT hardware often struggles to perform reliably under extreme temperatures, constant vibration, and dust exposure. To address these challenges, a leading autonomous vehicle operator in the mining sector sought a high-performance, industrial-grade server platform that could seamlessly integrate AI acceleration, ensure operational stability, and support mission-critical workloads in demanding conditions
System Requirements
The client required a robust edge server platform designed to handle the unique demands of autonomous mining vehicles. Key requirements included:
- Extreme environmental resilience: Reliable operation under wide temperature ranges and constant exposure to dust and debris.
- Vibration and shock resistance: Stability against continuous vehicle movement and sudden impacts, meeting industrial standards of vibration and shock.
- High throughput AI processing: Capability to run multiple deep learning models for object detection, terrain analysis, and safety monitoring in real time.
- Scalable expansion: Support for multiple PCIe Gen 5 accelerators to ensure computing flexibility for future AI workloads
- Advanced connectivity and management: High‑bandwidth networking with remote management for reliable fleet‑wide control and data synchronization.

System Description
To meet these stringent requirements, the client adopted a solution combining the Advantech ASMB-831 server board with the HPC-7420 4U rackmount chassis, delivering the ruggedness and computing power necessary for mission-critical autonomous operations. ASMB-831 Server Board: Powered by AMD® EPYC™ 9005/9004 series processors, the board supports up to six DDR5 RDIMM slots (576 GB) for high-speed memory performance. Its five PCIe Gen 5 x16 slots and dual 10 GbE Intel® X710-AT2 ports enable integration of multiple AI accelerators and rapid data transfer. With built-in IPMI, it allows remote diagnostics, monitoring, and power management across fleets deployed in remote mining sites.
HPC-7420 Industrial Chassis (Customized): Based on a rugged short-depth 4U enclosure, the chassis was custom-engineered to meet the client’s specific operational needs. It was upgraded to withstand 2 Grms vibration and 10 G shock, ensuring stability during heavy vehicle movement in rough mining terrain. Additionally, the chassis was equipped with a custom 24 V 2000 W DC power supply to integrate seamlessly with the autonomous vehicle’s onboard power system.
Together, the ASMB-831 and the customized HPC-7420 deliver a future-ready AI edge computing platform, enabling autonomous mining vehicles to perform real-time decision-making while maintaining operational safety and efficiency in one of the most demanding industrial sectors.
Benefits
- Reliable edge performance in mission-critical environments: Ensured continuous AI inference and data processing in extreme heat, cold, dust, and vibration.
- Reduced downtime and simplified maintenance: Front-access hot-swappable storage and robust chassis design improved serviceability in remote deployments.
- High throughput AI processing: Capability to run multiple deep learning models for object detection, terrain analysis, and safety monitoring in real time.
- High-speed communication: 10GbE ports enabled efficient data transmission from edge to central control systems.
- Remote fleet management: Built-in IPMI allowed centralized monitoring and control of distributed systems across mining sites.