ASUS IoT has unveiled its RUC-1000 series rugged-rack edge AI computers at COMPUTEX 2025, featuring 2 models of RUC-1000G and RUC-1000D, built to meet the demands of machine vision, smart factories, autonomous vehicles, and generative AI.
They host Intel Core Ultra 200S series processors and can do up to 64GB of DDR5-6400 RAM, with the RUC-1000G is positioned as the flagship, delivering up to 4000 AI TOPS with support for 600W GPUs via PCIe 5.0 – the first the brand has ever offered. It is also engineered for operation in harsh environments, with a temperature tolerance from -25°C to 60°C and an advanced thermal design that isolates the GPU chassis to maintain consistent performance.
In terms of design, the RUC-1000G adheres to a standardized 2U 19-inch rack-mount form factor, compatible with NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. It supports modular configurations and various mounting options, including wall and desk installations. Additionally, ASUS’s Q-Release Slim mechanism enables tool-free GPU removal, simplifying maintenance and minimizing mechanical risks.
Complementing the G model is the RUC-1000D, a compact, half-width fanless computer tailored for edge deployment as it features RAID 0/1/5/10 support and can house up to six 2.5-inch SSDs with an optional hot-swappable chassis. Connectivity includes 10GbE and 2.5GbE LAN, 10 USB ports, and six COM ports, along with multiple M.2 slots for 5G, WiFi 6, NVMe, and CAN bus modules. Its rugged build allows it to operate in even more extreme temperatures, from -25°C to 70°C.
Both models integrate security and remote management features aligned with industrial standards through compliance with IEC 62443-4-1 cybersecurity requirements, ensuring resilience against cyber threats, while the onboard iBMC module allows for out-of-band (OOB) monitoring, rapid system recovery, and reduced downtime. ASUS also equipped the units with MIL-STD-810H certification and wide-range 8-48V DC input with ignition control, reinforcing their suitability for demanding edge computing tasks.