Desktop, all-in-one and mini-PC platforms that run large mixture-of-experts models locally — up to Qwen3 235B — sourced in China, assembled and QA-tested in Saudi Arabia.
China AI hardware supply chain.
Saudi assembly, QA & integration.
Enterprise AI in production.
GCC commercialization channels.
The same 128GB unified-memory platform in a desk-side tower or a 27" all-in-one — enough headroom to run a 235B-parameter mixture-of-experts model locally.
Both platforms ship with Windows 11 / Ubuntu 24.04 and support local deployment of Qwen3 235B (mixture-of-experts).
A pocket-sized AMD platform for edge deployment — desks, kiosks and field stations that don't have room for a tower.
Supports local deployment of Qwen3 30B (mixture-of-experts) · Windows 11 / Ubuntu 24.04
A 500-gram, 128×134×52mm chassis available with either Intel's latest efficiency-class or performance-class silicon.
Windows 11 / Linux · ~500 g · 128 × 134 × 52 mm · adapter 90–120 W depending on variant
A large quadruped robot with twelve degrees of freedom, dual onboard computers, and full-surround perception — stable on rough terrain, stairs, and in total darkness.
