Anshtern Debuts Full Stack Intelligent Computing and Embodied Intelligence Engine

Anshtern made its first official appearance at the Magic X event in Silicon Valley, unveiling full stack intelligent computing solutions and a general embodied intelligence engine for global partners and industry leaders, announced in a press release.

Vice President Troy Shen introduced the new engine, designed to address data scarcity, reliable complex robotic operations, and high computing costs. The system is built on four core elements: the Data Flywheel, Long Horizon Embodied Brain, Robotic Data Factory, and Dedicated AI Computing Power. Together these components support a cloud edge device integrated open platform for scalable industrial deployment.

At the event, Anshtern presented three main business segments. The first is high performance AI computing infrastructure, including bare metal servers, AI large model clusters, and the proprietary KLLM inference engine. The second is the Shanghai Songjiang Intelligent Computing Center, which provides 1000P total computing power with a PUE below 1.2 and 99.999 percent service availability. The third is an Embodied AI Data Training Center combining physical and simulation environments to support humanoid robot development.

Anshtern stated that it plans to continue expanding its global collaborations and focus on embodied AI innovation through its intelligent computing ecosystem.