With the continuous improvement of supercomputer performance and the integration of artificial intelligence with traditional scientific computing,the scale of applications is gradually increasing,from millions to tens...With the continuous improvement of supercomputer performance and the integration of artificial intelligence with traditional scientific computing,the scale of applications is gradually increasing,from millions to tens of millions of computing cores,which raises great challenges to achieve high scalability and efficiency of parallel applications on super-large-scale systems.Taking the Sunway exascale prototype system as an example,in this paper we first analyze the challenges of high scalability and high efficiency for parallel applications in the exascale era.To overcome these challenges,the optimization technologies used in the parallel supporting environment software on the Sunway exascale prototype system are highlighted,including the parallel operating system,input/output(I/O)optimization technology,ultra-large-scale parallel debugging technology,10-million-core parallel algorithm,and mixed-precision method.Parallel operating systems and I/O optimization technology mainly support largescale system scaling,while the ultra-large-scale parallel debugging technology,10-million-core parallel algorithm,and mixed-precision method mainly enhance the efficiency of large-scale applications.Finally,the contributions to various applications running on the Sunway exascale prototype system are introduced,verifying the effectiveness of the parallel supporting environment design.展开更多
Correct extraction of the ultra-large-scale integrated (ULSI) interconnect components at hight frequencies is very important for evaluating electrical performances of high-speed ULSI circuits.In this paper, the extrac...Correct extraction of the ultra-large-scale integrated (ULSI) interconnect components at hight frequencies is very important for evaluating electrical performances of high-speed ULSI circuits.In this paper, the extraction of the interconnect resistance at high frequencies is derived from the Ohm′s law and verified by the software FastHenry.The results are also compared with those of another resistance formula originated from the effective area of the current flowing. The applicability of these two formulae is discussed.The influence of the interconnect geometry on the resistance at high frequencies is studied.The computation indicates that the effect of frequency on the resistance is weak when the skin depth is larger than half of the short side of the rectangular interconnect cross section.With further increase of frequency, the resistance increases obviously. Results imply that conductor with a square cross section exhibits the largest resistance for rectangular conductors of constant cross section area.展开更多
基金Project supported by the Key R&D Program of Zhejiang Province,China(No.2022C01250)the National Key R&D Program of China(No.2019YFA0709402)。
文摘With the continuous improvement of supercomputer performance and the integration of artificial intelligence with traditional scientific computing,the scale of applications is gradually increasing,from millions to tens of millions of computing cores,which raises great challenges to achieve high scalability and efficiency of parallel applications on super-large-scale systems.Taking the Sunway exascale prototype system as an example,in this paper we first analyze the challenges of high scalability and high efficiency for parallel applications in the exascale era.To overcome these challenges,the optimization technologies used in the parallel supporting environment software on the Sunway exascale prototype system are highlighted,including the parallel operating system,input/output(I/O)optimization technology,ultra-large-scale parallel debugging technology,10-million-core parallel algorithm,and mixed-precision method.Parallel operating systems and I/O optimization technology mainly support largescale system scaling,while the ultra-large-scale parallel debugging technology,10-million-core parallel algorithm,and mixed-precision method mainly enhance the efficiency of large-scale applications.Finally,the contributions to various applications running on the Sunway exascale prototype system are introduced,verifying the effectiveness of the parallel supporting environment design.
文摘Correct extraction of the ultra-large-scale integrated (ULSI) interconnect components at hight frequencies is very important for evaluating electrical performances of high-speed ULSI circuits.In this paper, the extraction of the interconnect resistance at high frequencies is derived from the Ohm′s law and verified by the software FastHenry.The results are also compared with those of another resistance formula originated from the effective area of the current flowing. The applicability of these two formulae is discussed.The influence of the interconnect geometry on the resistance at high frequencies is studied.The computation indicates that the effect of frequency on the resistance is weak when the skin depth is larger than half of the short side of the rectangular interconnect cross section.With further increase of frequency, the resistance increases obviously. Results imply that conductor with a square cross section exhibits the largest resistance for rectangular conductors of constant cross section area.
