the best performance and parallelization of hardware and software, but these technologies have matured. In terms of technological development, high-speed computing has advanced to Container and Container Orchestration, which enable various industries to operate smoothly in different workflows. Dr. Zhang Youcheng pointed out that this is still not the best solution for enterprises to upgrade their industries or improve their processes.

At present, the biggest problem facing industrial upgrading phone database is how to customize the development of applications at the application layer for industries in various fields under a sound software and hardware architecture. The important value of high-speed computing should be to connect computing power technology with the needs of end users. Taking medical imaging AI interpretation as an example, if a doctor wishes to analyze the patient's pathological image data to optimize the medical direction, in terms of the current practical situation, a complete information team must first be required to handle the software and hardware required for high-speed computing.

Physical problems, but not every doctor can have such team support, which means that the computing power of HPC is not really passed on to those who want to use it. "If the application of high-speed computing technology cannot be applied to end users, it will be difficult to improve medical quality and digital transformation." High-speed computing technology must also keep pace with the times. Dr. Zhang Youcheng explained that, in terms of AI application, the operation of pure numbers is quite simple, but with the evolution of demand, there are not only more images, but also the trend of "Federated Learning".