Research on Design of High-Performance Graph Computing System Based on CPU Heterogeneous Architecture
Yanjiang Li · 2021 International Conference on Electronic Information Technology and Smart Agriculture (ICEITSA) · 2021
High performance computing is also known as supercomputing, mainly refers to more data in a short time to help the computer system to complete more computing tasks. Cloud computing, as a parallel distributed computing application concept proposed by the current technological innovation, will use the network to connect a large number of computing resources in practical application, thus forming a computing resource pool with stronger computing capacity and parallel processing capacity, so as to bring more superior computing services to users. From the perspective of overall development, cloud computing, as a new processing mode proposed to achieve high performance computing, is very novel in both application scope and promotion form, which plays a positive role in the future research of computer technology. Therefore, based on the understanding of high performance computing research significance and the central processing unit (CPU) implementation on the basis of the advantages of high performance computing, start from a computer unified device architecture (CUDA) for fast image processing system, in-depth studies are proving CPU heterogeneous architecture as the core of high performance computing system has a certain application advantages.