Performance Analysis and Science Mapping on High Performance Computing in the Era of Artificial Intelligence
摘要
The demand for high-performance computing (HPC) systems is increasing due to the growing need for large-scale computation and the rapid evolution of artificial intelligence. Previous bibliometric analyses on HPC are outdated, leaving the research community without an up-to-date investigation of the field. To bridge this gap and support the development of HPC while informing policymakers, this paper conducts a performance analysis and science mapping of HPC research. The study examines different aspects, including publication characteristics, research disciplines, trends, keyword co-occurrence, prolific authors, contributing countries, funding agencies, and document types. Findings indicate that researchers from the United States of America lead in HPC research. The Oak Ridge National Laboratory contributed the highest number of articles in the field of HPC. It is found that HPC research results are predominantly published in conferences. The top two leading funders of HPC research are from the United States of America. The United States of America play a pivotal role in linking different regions, facilitating cross-border research and innovation in HPC research. Mapping analysis shows that High-Performance Computing, Cloud Computing, Scientific Computing and Sparse Matrix-Vector Multiplication as frequent keywords highlights the core focus areas in HPC research. The article serves as a valuable resource for the research community, policymakers, funding agencies, industry professionals, and students seeking insights into the landscape of HPC research in this era of AI.