MAXER-2100 – 2U Rackmount AI Inference Server with NVIDIA RTX 4080 Super by AAEON

The AAEON MAXER-2100 is a purpose-built 2U rackmount AI inference server designed for engineers and system integrators deploying demanding machine learning inference, computer vision, industrial inspection, and GPU-accelerated analytics workloads in data centre and edge rack environments. Built on the Intel Q670 chipset with 12th and 13th Generation Intel Core LGA1700 socket processors, the standard configuration ships with an Intel Core i9-13900 and supports up to the i9-13900K at 125W TDP, providing the CPU throughput needed to sustain high-performance GPU inference pipelines without processor bottlenecks. The integrated NVIDIA GeForce RTX 4080 Super in a PCIe x16 slot can optionally be replaced with other NVIDIA GeForce or dedicated AI computing cards including the RTX 4090, giving users flexibility to scale GPU compute to their specific inference requirements. System memory is provided by four DDR5 4000MHz DIMM slots supporting up to 128GB in dual-channel non-ECC configuration, and storage is covered by two hot-swappable 2.5 inch SATA-III drive bays with RAID 0 and RAID 1 support plus one M.2 2280 NVMe slot for the operating system. Networking includes one Intel AMT 12.0-enabled GbE management port, one additional GbE, and two 2.5GbE ports for high-bandwidth data ingestion. Four USB 3.2 Gen 2 ports at 10Gbps, one RS-232/422/485 serial port, and a front access I/O design round out the connectivity. Onboard TPM 2.0 covers platform security. The built-in 1000W power supply handles the full system thermal envelope. Operating temperature is 0°C to 40°C with 0.5m/s airflow. CE and FCC Class A certified. Available from Recab, your Nordic partner for AAEON AI inference server solutions.