Max Robotics — Global Service Robot Directory + US Certification

About

Shengwen Liang is a leading researcher in efficient AI computing, with a focus on deploying large-scale models on resource-constrained edge devices. His work bridges the gap between massive deep learning models and practical hardware, addressing critical challenges in on-device intelligence. Liang's most notable contribution is the "Cambricon-LLM: A Chiplet-Based Hybrid Architecture for On-Device Inference of 70B LLM" (2024, 17 citations), which pioneers a novel chiplet design to run 70-billion-parameter language models on smartphones and robotics—a feat previously thought impossible due to memory and power limits. This work tackles single-batch, low-arithmetic-intensity computing, enabling privacy-preserving AI without cloud dependency. Liang has also contributed to the broader edge AI community through his involvement in the "2020 Low-Power Computer Vision Challenge" (2021, 3 citations), which benchmarks energy-efficient vision systems for IoT and battery-powered devices like drones. His research directly impacts the future of autonomous systems, mobile AI, and privacy-centric computing, making him a key figure in the push toward practical, on-device intelligence.