Albert Liu

Albert Liu is founder and chief executive of Kneron, a company he started in 2015 to design neural processing units (NPUs), chips built to run AI models on devices and local servers. He holds a PhD in electrical engineering from UCLA and worked on AI hardware and chip systems at Qualcomm, Samsung Electronics’ R&D center, MStar and Wireless Information before Kneron, according to a SemiWiki interview.

Reconfigurable accelerators

Liu holds more than 30 international patents and has published more than 70 technical papers. He received the IEEE Darlington Award for work on reconfigurable AI accelerator architectures, chips that can be rearranged for different neural networks, and he wrote the book Artificial Intelligence Hardware Design: Challenges and Solutions.

Inference first

Kneron was built on the belief that inference, running trained models, would become the main AI workload. “Training models is no longer the primary challenge,” Liu told SemiWiki in 2026. He says the company’s advantage is time: “Many companies can build accelerators. Far fewer companies have spent nearly a decade building infrastructure specifically optimized for real-world inference deployment.” Kneron now sells chips and edge servers for security systems, cars and AI PCs, where power use, latency and privacy decide whether AI can run locally.