ASUS has introduced the RUC-2000 series rugged-rack edge AI system, led by the fanless 2U RUC-2000H, at COMPUTEX 2026. The platform targets space-constrained industrial deployments running edge inference workloads such as machine vision and video analytics, with ASUS listing performance up to 180 AI TOPS on an Intel Core Ultra Series 3 processor.
The mechanical hook is a modular half-rack format. The RUC-2000H is a 2U half-rack base unit, and two units can be combined into a standard 19-inch rack configuration that ASUS calls RUC-2000G. ASUS also calls out mounting flexibility (rack, wall, or desk) aimed at simplifying installation and service across mixed sites.
On the I/O side, the RUC-2000 series includes one 10GbE port and five 2.5GbE ports. For camera-heavy edge deployments, the system supports up to eight GMSL2 interfaces for multi-camera aggregation. Storage is built around dual hot-swappable SSD trays with RAID 0/1 support, plus additional hot-swappable storage options intended to speed recovery and reduce downtime.
For expansion, ASUS lists a PCIe x16 slot wired for x8 bandwidth, with support for up to a 200 W GPU. That matters for edge sites that start with CPU-based inference but later need to add accelerators without swapping the entire platform.
But the key engineering bet here is environmental hardening without active airflow. ASUS says the RUC-2000H uses a fanless architecture to reduce dust-related failure risk, supports wide-range DC input for power protection, and is validated to MIL-STD-810H rugged standards for industrial environments with temperature extremes, dust, and vibration. Fanless edge systems can be a practical fit for dirty or maintenance-limited locations, but they also put more pressure on enclosure thermal design, placement, and any local conduction or convection path you can realistically provide.
ASUS says RUC-2000 series samples are available now through local ASUS representatives.
Source: ASUS
