Report Title: Large Models and Large-Scale Computing

Speaker: Dr. Jianguang Lou

Affiliation: SUNMMIO Technology Ltd (Beijing)

Organizer: School of Computer Science and Information Engineering

Time: 15:00–17:00, Thursday, October 30, 2025

Venue: Room B501, Feicui Science and Technology Building, Feicui Lake Campus

Report Abstract:

The release of ChatGPT at the end of 2022 was a groundbreaking event, making the public see the hope of artificial general intelligence for the first time. However, why can a language model trained with a very simple training objective based on Transformer, a generative artificial neural network, ignite the dawn of artificial general intelligence? In this report, we will explore the essence of intelligence together, and from the perspective of the essence of intelligence, understand the progress of large models in general intelligence in recent years, and look forward to the next development directions and research hotspots of large models. At the same time, the development of large models has brought huge computing demands, posing new challenges to basic hardware and software stacks. This report will explore the opportunities for hardware-software co-design from the perspective of the computing mode of large models.

Speaker Profile:

Jianguang Lou, Ph.D., is currently the Chief AI Scientist of SUNMMIO Technology Ltd (Beijing), responsible for leading the company's AI software ecosystem team to provide support for advanced AI chips. Before joining SUNMMIO Technology in September 2025, he served as a Senior Principal Research Manager at Microsoft Research, leading a team of researchers and engineers focusing on cutting-edge AI-related research projects, including research on interactive data analysis robots, automatic data understanding, automatic program generation, intelligent diagnosis and operation and maintenance of large-scale online systems, and applying the research results to major Microsoft products. He has about 20 years of research and application experience in the fields of machine learning, natural language processing, artificial intelligence and AI for software systems, and has published more than 150 research papers in top academic conferences.