Presentation given at AIChE Computing and Systems Technology Awards Banquet, November 17, 1998, Miami Beach, Florida
Mark A. Stadtherr
Department of Chemical Engineering
University of Notre Dame
Notre Dame, IN 46556 USA
In the first part of this presentation we will take a brief look at the tremendous growth in computational power that is ongoing, from desktop machines to high performance computing hardware. It will come as no surprise to this audience that computational power has grown very quickly and continues to grow very quickly. The availability of all this power allows us to solve problems much faster. This means we can solve larger problems involving more complex and more realistic models. It allows us to solve problems we probably would not have even considered trying to solve years ago, because of their computational demands. All of this has made possible significant advances in many fields of science and engineering. But, another thing we can use all this computer power for, that is often overlooked in the quest to solve problems faster and faster, is that we can use it to solve problems more reliably—in fact to actually provide mathematical and computational guarantees of reliability. So in the second part of this presentation we will focus on issues of reliability—what good is high performance computing if we are just computing the wrong answer faster?