We develop GPGPU chips and intelligent computing solutions based on GPGPU to provide the basic computing capabilities required for AI. By integrating self-developed GPGPU-based hardware and the proprietary BIRENSUPA software platform, our solutions support the training and inference of AI models in a wide range of applications from the cloud to the edge. In particular, our GPGPU-based solutions have strong performance and high efficiency in pre-training, post-training, and reasoning for large language models (“LLMs”), and have high technical barriers, giving us a key advantage in domestic competition. Our technology is an important infrastructure that supports the development of AI and promotes the progress of AGI. It can meet the growing computing needs of various Industry, thereby helping to improve productivity, innovation and transformation.