Form： 爱集微 2018/11/22 Browse：1680 Keywords: Ying Wei Da
Huang Renxun said that AI, as one of the most powerful technologies ever, enables machines to learn from large amounts of data and write software. With AI, industries can learn from massive data to build prediction models, and apply these models to the industry environment or operation.
The more data, the stronger the ability to train powerful AI models. E-commerce, retail, financial services, telecommunications, health care are becoming data-driven AI industries.
In Huang Renxun's view, AI is a new computing method, which is changing the pattern of computing field. In large computing field, HPC and super large data center are running at a very high speed.
HPC is becoming a AI computer and used by scientists, data scientists and AI developers. Scientists are integrating physical modeling and AI forecasting methods to create larger orders of magnitude. This attempt was implemented with the support of multi precision Volta Tensor Core GPU.
Meanwhile, data scientists and AI developers need to use HPC to train more complex AI models on more data.
Internet Co relies heavily on machine learning to create personalized services that provide personalized recommendation. Very large-scale nodes must act as a HPC cluster to train models or serve millions of concurrent users running various AI models.
Huang Renxun said that super-large-scale computing is now also HPC. As a cloud service, large-scale HPC will attract scientists and industry people to use it. Because of AI, large-scale computing and enterprise computing will join scientific computing and become a new HPC.