Last week, the 9th WU Wen-Jun AI Science and Technology Awards were announced. Prof. YU Hao (School of Microelectronics) and his team received an AI Technology Invention Award for a project entitled “Development of Technology for Deep Learning Chips towards Edge Computing.” Southern University of Science and Technology (SUSTech) is the first-author affiliation for this award.
The WU Wen-Jun AI Science and Technology Awards are regarded as the highest award for artificial intelligence science and technology in China, which is sponsored by the Chinese Association for Artificial Intelligence (CAAI). The WU Wen-Jun AI Science and Technology Award is named after Prof. WU Wen-Jun, an artificial intelligence pioneer scholar who is one of the only two recipients of the inaugural State Preeminent Science and Technology Award.
Prof. YU Hao and his team have made it possible for efficiently mapping deep learning algorithms on edge devices with limited computing resources. This team has been recognized international recognition on energy-efficient and high-throughput AI algorithm and chip design. They have developed trained quantization algorithms for deep convolutional neural networks (CNN), as well as tensor compression algorithms for deep recurrent neural networks (RNN), which can be further realized as high-throughput and low-power edge AI chips. As such, it leads to a hardware with a performance similar to the existing GPU chips but with dramatically reduced power and cost. Such a research result shows differentiation in AI technology, and is in-line with the needs for upgrading Chinese high-tech enterprises for the coming edge-AI era, especially in Shenzhen.
