Maxsun has built a computer that packs serious punch into a tiny package. The MoDT dual-GPU Mini Station crams desktop power into a box smaller than most gaming rigs. Tech teams working with artificial intelligence and data science get everything they need without taking up half their office space. The machine runs Intel's latest Core Ultra 9 285HX processor alongside two powerful Ruixuan Pro B60 graphics cards. Each graphics card brings 24GB of memory for a total of 48GB that can handle massive AI models like Qwen3-32B.
The company solved the heat problem that usually kills compact computers. Dual fans work with special heat pipes and metal plates to keep temperatures under control during heavy workloads. Workers can run intensive AI training sessions for hours without the system slowing down from overheating. The whole setup stays quiet enough for regular office use while delivering performance that rivals much larger machines. Three different storage slots let users add lightning-fast drives for their data.
Maxsun loaded the Mini Station with connection options that make expansion easy. Four Thunderbolt ports provide blazing-fast links to external devices and storage arrays. The system comes ready to run with either Windows or Linux operating systems. Engineers and researchers can deploy the machine for on-site AI work or local model training without needing a dedicated server room.
The company solved the heat problem that usually kills compact computers. Dual fans work with special heat pipes and metal plates to keep temperatures under control during heavy workloads. Workers can run intensive AI training sessions for hours without the system slowing down from overheating. The whole setup stays quiet enough for regular office use while delivering performance that rivals much larger machines. Three different storage slots let users add lightning-fast drives for their data.
Maxsun loaded the Mini Station with connection options that make expansion easy. Four Thunderbolt ports provide blazing-fast links to external devices and storage arrays. The system comes ready to run with either Windows or Linux operating systems. Engineers and researchers can deploy the machine for on-site AI work or local model training without needing a dedicated server room.
