摘要
科学计算的兴起是20世纪后半叶最重要的科技进步之一。计算与理论及实验相并列,已经成为当今世界科学活动的第三种手段。为把信息和数据变成知识,从而探索科学术知,促进技术创新,加强国防建设,保障国家安全,计算将起不可替代的重要作用。当前科学计算正在向大规模和高性能发展,要达到“全物理、全系统、三维、高分辨、高逼真”的数值模拟,发展高效的计算方法与发展高性能的计算机同等重要。大规模计算提出的世界性难题已形成科学计算的学科前沿。求解由实际问题得到的复杂的偏微分方程不仅计算规模大,更由于非线性、多尺度、长时间、不适定、多区域、高病态等特点使计算格外困难。现有的算法远不能满足需求,这正是本项目必须解决的关键科学问题。本项目是多学科交叉的应用基础研究,将以针对国家目标在环境、材料、能源等领域选择的几个挑战性的大规模计算问题为主攻方向,应用科学计算这一研究手段,为高新技术及前沿学科的发展提供必要的数据和新的研究途径。本项目将重点研究解决这些重大问题所必需的高性能计算方法,并利用我国已有和将发展的大型计算机系统为测试和实现平台,解决几个有实际背景的大规模科学计算问题。本项目设置了复杂流动的高精度数值模拟,物质性质机理的多尺度计算研究。
Fast advances of information technology are making scientific computing as valuable as theory and experimentation as a tool for sientific research. New supercomputers require fundamentally new approaches to scientific computation, especially, effective use of such machines requires separating problem into pieces to be worked on by different small computers and then reassembling the pieces. It is very difficult to do in practice. This project is suggested for basic research with some definite applied objectives, which are challenging to computing on large scales in the fields of environment, material and energy, encloding the computation of complex flows with high accuracy, the computation for material properties with multiscales, the numerical simulation of oil storages and the computation of wave problems. Some new computational methods and algorithms with high performance, which break through some natural defences caused by the nonlinear properties, multiscales, and long-term computations, will be created and developed, some applicable programs will be given. By using the available advanced computer systems, especially, made in China, as the platform for test and implementation of computation, some important practical computing problems mentioned above will be solved, while an important place in the front of scientific computing in the world will be won.
出处
《中国基础科学》
2001年第1期19-25,共7页
China Basic Science
