A new paradigm is emerging in mechanics. While the historical development of our industrialized world owes its greatest debt to the analytical pursuit of mechanics in the Newtonian tradition, the limits of such efforts are becoming very clear. The human race has now entered a stage of sophistication where it is willing and ready to address problems that cannot be attacked within the simplifications required in purely analytical modeling. Problems in turbulence and acoustics, complex materials, biological systems, and countless others, as well as problems reaching beyond the boundaries of those fields, will need more than simple formula to be described accurately.
Such problems involve large numbers of different elements interacting in complex ways. While the average result of all the interactions may often appear relatively simple, predicting this average behavior reliably is perhaps the biggest problem facing us today.
Yet, along with this problem, a tool for its resolution has also emerged. Modern computers have achieved a power that allows, (or will within a foreseeable time allow), the solution of many of such complex systems in very realistic situations. Progress beyond the realm of mechanics previously accessible depends on harnessing the power of computers.
This is not a trivial endeavor. Not only do the problems we face often strain our most optimistic visions of the future of computing; when we are successful, we need to learn to use the fruits of our computational explorations. An incredible amount of raw results awaits us; we must learn to identify important similarities and important differences. After computing the detailed behavior of the may be millions of elements forming a complex system, we still need to find out what are the key overall properties of that system. Only then can we predict how our system interacts within its own, larger environment.
It is a formidable challenge to effectively apply the capabilities of massively parallel and distributed supercomputers, and to learn from their results. It is a challenge that requires a new thinking, a commitment to using the new tools, and a willingness to go new directions.