Taipei, Taiwan – August 11, 2026 – Neousys Technology (TWSE: 6922), a leading provider of rugged embedded systems, will exhibit at the 2026 Taipei International Industrial Automation Exhibition (Automation Taipei) from August 19 to 22 at Booth M1008. Themed “Rugged Edge, Real-Time AI Computing,” Neousys will highlight practical applications of Generative AI and Physical AI in industrial environments, delivering real-time compute performance and reliability in harsh environments. Through four live demonstrations, visitors will gain comprehensive insights into how next-generation smart factories and semiconductor supply chains are putting edge AI to work.
Driven by advances in Generative AI and increasingly autonomous systems, achieving low-latency, resilient edge computing in demanding factory environments has become pivotal to successful industrial transformation. As an edge computing pioneer, Neousys addresses three critical manufacturing pain points: high-precision semiconductor inspection, smart workplace safety, and mobile platform autonomy. In collaboration with partners LEYU Precision, IQUMIN Vision, and Nineeon, Neousys will present four live demos illustrating how rugged edge platforms power reliable AI deployments across a wide range of industrial applications.
Beyond industrial applications, Neousys will unveil three next-generation rugged computing platforms tailored for diverse edge workloads - ranging from lightweight AI inference to extreme high-performance computing (HPC) in harsh environments:
- POC-900/960GC Series Rugged Compact Computers: Powered by the x86-based AMD® Ryzen™ PRO 8640U processor, the POC-900/960GC series strikes an ideal balance between a compact form factor and outstanding performance. Featuring Ryzen™ AI with an integrated 16 TOPS NPU accelerator, it is purpose-built for space-constrained deployments - such as small control cabinets, narrow machine enclosures, and real-time AI vision processing - delivering a low-power, high-efficiency edge AI computing solution.
- Jemii-150 Series Mini-ITX NVIDIA Jetson Carrier Boards: Built on a standard Mini-ITX form factor, the Jemii-150 series fully supports NVIDIA Jetson Orin NX and Orin Nano modules and is compatible with all NVIDIA JetPack versions. Delivering up to 157 TOPS of powerful AI inference in a compact profile, it is fully compatible with NVIDIA open models, Jetson software stack and agent skills, helping developers significantly cut time-to-market and reduce deployment risks to edge devices.
- RGS-9805GC Series Rugged HPC Servers: Powered by the latest AMD® EPYC™ 9005 Turin processors, the RGS-9805GC brings data center-class computing to rugged edge environments for the first time. Supporting up to two NVIDIA RTX PRO™ 6000 Blackwell Max-Q Workstation EditionGPUs, it delivers massive CPU/GPU heterogeneous computing performance. Engineered to withstand extreme temperatures, severe shock, and heavy vibration, this platform is purpose-built for "Sense-Think-Act" Physical AI applications - including high-speed autonomous trucking, in-vehicle HD mapping, and real-time autonomous logistics - ensuring 24/7 mission-critical uptime.
"As Physical AI accelerates the interaction and adaptation of autonomous robots within real-world environments, enabling automated systems to learn and adapt in real time, AI deployment at the edge will continue to advance at a rapid pace," said Bryan Lan, Vice President of Neousys Technology. "From lightweight AI computing platforms to rugged high-performance industrial servers, our three new product families unveiled this year are designed to provide customers with a comprehensive range of computing solutions, helping industries break through environmental limitations and accelerate their AI transformation."
Neousys will present a wide range of live demonstrations and application showcases at the exhibition, where its technical team will be available on-site to answer application-related questions. We sincerely invite industry professionals to visit our booth to explore the limitless possibilities of AI-driven automation together.