About performance and intellectuality of supercomputer modeling


如何引用文章

全文:

开放存取 开放存取
受限制的访问 ##reader.subscriptionAccessGranted##
受限制的访问 订阅存取

详细

The concept of supercomputer technologies is traditionally related to mapping algorithms onto the computer architecture, which, taking into account the explosive growth of computational capabilities, implies the necessity for an adequate increase in the performance of algorithms and programs. At the same time, it is well known that the rate of building “computer muscles” far exceeds the rate of increasing the labor productivity of software developers, which becomes a bottleneck of computer evolution. The only way to deal with this problem is to automate the construction of models, algorithms, and programs, which directly implies the revolutionary change in the level of artificial intelligence in supercomputer technologies. In this paper, it is from this standpoint that main computational stages of mathematical modeling of various processes and phenomena are discussed, some aspects of high logical complexity of modern high-performance methods for solving “large” applied problems are pointed out, and some intelligent solutions for various modeling problems are proposed.

作者简介

V. Il’in

Institute of Computational Mathematics and Mathematical Geophysics, Siberian Branch; Novosibirsk State University

Email: iskopin@gmail.com
俄罗斯联邦, pr. Akademika Lavrent’eva 6, Novosibirsk, 630090; ul. Pirogova 2, Novosibirsk, 630090

I. Skopin

Institute of Computational Mathematics and Mathematical Geophysics, Siberian Branch; Novosibirsk State University

编辑信件的主要联系方式.
Email: iskopin@gmail.com
俄罗斯联邦, pr. Akademika Lavrent’eva 6, Novosibirsk, 630090; ul. Pirogova 2, Novosibirsk, 630090

补充文件

附件文件
动作
1. JATS XML

版权所有 © Pleiades Publishing, Ltd., 2016