ASUS is expanding its AI strategy through a new Agent Computer initiative designed to bring local and agentic AI to a variety of desktop hardware. The company stated that users can choose between local, cloud, or hybrid AI models based on their specific performance, privacy, and cost requirements. This ecosystem spans DIY PCs, gaming systems, workstations, and mini PCs, utilizing motherboards from the PRIME, TUF Gaming, ROG, and ProArt families.
The initiative includes hardware such as the ExpertCenter Pro ET900A X9 workstation for demanding workloads and the ProArt GR1X mini PC. The mini PC utilizes NVIDIA's RTX Spark platform, featuring a 6,144-core Blackwell GPU, a 20-core Grace CPU, and up to 128 GB of unified memory. This configuration allows the device to run large AI models and agentic workflows locally, including models with up to 120 billion parameters.
To support mobile users and laptops, ASUS has launched the XG Core external GPU. This device is powered by a low-profile Radeon RX 9060 XT with 16 GB of VRAM and is intended to provide local AI acceleration alongside high-performance gaming. The company also introduced a "Find Your ASUS Agent Computer" tool on its official website to help users select components based on their specific AI needs.
Software development is a central component of the ecosystem. ASUS is developing Zenni Claw to bring agentic AI capabilities to consumer devices and Multi-LM Tuner to assist users in deploying optimized large language models across its platforms. Additionally, the company is deploying DisplayWidget Center to serve as a bridge between ASUS displays and emerging AI workflows.
